diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8440 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999459527627146,
+  "eval_steps": 100,
+  "global_step": 3469,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008647557965661989,
+      "grad_norm": 10.80162525177002,
+      "learning_rate": 1.7241379310344828e-07,
+      "loss": 2.2563,
+      "step": 3
+    },
+    {
+      "epoch": 0.0017295115931323978,
+      "grad_norm": 9.61453914642334,
+      "learning_rate": 3.4482758620689656e-07,
+      "loss": 2.0897,
+      "step": 6
+    },
+    {
+      "epoch": 0.0025942673896985967,
+      "grad_norm": 9.470385551452637,
+      "learning_rate": 5.172413793103449e-07,
+      "loss": 1.9261,
+      "step": 9
+    },
+    {
+      "epoch": 0.0034590231862647956,
+      "grad_norm": 9.873086929321289,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 2.0147,
+      "step": 12
+    },
+    {
+      "epoch": 0.004323778982830994,
+      "grad_norm": 10.191353797912598,
+      "learning_rate": 8.620689655172415e-07,
+      "loss": 1.858,
+      "step": 15
+    },
+    {
+      "epoch": 0.005188534779397193,
+      "grad_norm": 10.490758895874023,
+      "learning_rate": 1.0344827586206898e-06,
+      "loss": 2.2373,
+      "step": 18
+    },
+    {
+      "epoch": 0.006053290575963392,
+      "grad_norm": 10.7379789352417,
+      "learning_rate": 1.2068965517241381e-06,
+      "loss": 2.2257,
+      "step": 21
+    },
+    {
+      "epoch": 0.006918046372529591,
+      "grad_norm": 9.31098747253418,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 1.9876,
+      "step": 24
+    },
+    {
+      "epoch": 0.00778280216909579,
+      "grad_norm": 9.79057502746582,
+      "learning_rate": 1.5517241379310346e-06,
+      "loss": 1.9015,
+      "step": 27
+    },
+    {
+      "epoch": 0.008647557965661988,
+      "grad_norm": 9.3207368850708,
+      "learning_rate": 1.724137931034483e-06,
+      "loss": 1.8092,
+      "step": 30
+    },
+    {
+      "epoch": 0.009512313762228188,
+      "grad_norm": 8.36523151397705,
+      "learning_rate": 1.896551724137931e-06,
+      "loss": 1.7048,
+      "step": 33
+    },
+    {
+      "epoch": 0.010377069558794387,
+      "grad_norm": 7.840041160583496,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 1.4938,
+      "step": 36
+    },
+    {
+      "epoch": 0.011241825355360585,
+      "grad_norm": 8.646334648132324,
+      "learning_rate": 2.241379310344828e-06,
+      "loss": 1.4769,
+      "step": 39
+    },
+    {
+      "epoch": 0.012106581151926784,
+      "grad_norm": 8.217284202575684,
+      "learning_rate": 2.4137931034482762e-06,
+      "loss": 1.4442,
+      "step": 42
+    },
+    {
+      "epoch": 0.012971336948492982,
+      "grad_norm": 8.16273307800293,
+      "learning_rate": 2.5862068965517246e-06,
+      "loss": 1.6091,
+      "step": 45
+    },
+    {
+      "epoch": 0.013836092745059182,
+      "grad_norm": 7.606343746185303,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 1.565,
+      "step": 48
+    },
+    {
+      "epoch": 0.01470084854162538,
+      "grad_norm": 5.831145763397217,
+      "learning_rate": 2.931034482758621e-06,
+      "loss": 1.4604,
+      "step": 51
+    },
+    {
+      "epoch": 0.01556560433819158,
+      "grad_norm": 5.626949310302734,
+      "learning_rate": 3.103448275862069e-06,
+      "loss": 1.2422,
+      "step": 54
+    },
+    {
+      "epoch": 0.01643036013475778,
+      "grad_norm": 8.196585655212402,
+      "learning_rate": 3.2758620689655175e-06,
+      "loss": 1.178,
+      "step": 57
+    },
+    {
+      "epoch": 0.017295115931323976,
+      "grad_norm": 3.9342713356018066,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 1.1149,
+      "step": 60
+    },
+    {
+      "epoch": 0.018159871727890176,
+      "grad_norm": 5.660026550292969,
+      "learning_rate": 3.620689655172414e-06,
+      "loss": 1.2666,
+      "step": 63
+    },
+    {
+      "epoch": 0.019024627524456376,
+      "grad_norm": 3.853914737701416,
+      "learning_rate": 3.793103448275862e-06,
+      "loss": 0.9817,
+      "step": 66
+    },
+    {
+      "epoch": 0.019889383321022573,
+      "grad_norm": 4.146341323852539,
+      "learning_rate": 3.96551724137931e-06,
+      "loss": 1.0925,
+      "step": 69
+    },
+    {
+      "epoch": 0.020754139117588773,
+      "grad_norm": 3.1577234268188477,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 0.8252,
+      "step": 72
+    },
+    {
+      "epoch": 0.02161889491415497,
+      "grad_norm": 3.2642972469329834,
+      "learning_rate": 4.310344827586207e-06,
+      "loss": 0.7533,
+      "step": 75
+    },
+    {
+      "epoch": 0.02248365071072117,
+      "grad_norm": 3.2303547859191895,
+      "learning_rate": 4.482758620689656e-06,
+      "loss": 0.7326,
+      "step": 78
+    },
+    {
+      "epoch": 0.02334840650728737,
+      "grad_norm": 2.59647536277771,
+      "learning_rate": 4.655172413793104e-06,
+      "loss": 0.7174,
+      "step": 81
+    },
+    {
+      "epoch": 0.024213162303853567,
+      "grad_norm": 2.8148062229156494,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 0.7488,
+      "step": 84
+    },
+    {
+      "epoch": 0.025077918100419767,
+      "grad_norm": 1.758013129234314,
+      "learning_rate": 5e-06,
+      "loss": 0.7332,
+      "step": 87
+    },
+    {
+      "epoch": 0.025942673896985964,
+      "grad_norm": 2.0942482948303223,
+      "learning_rate": 5.172413793103449e-06,
+      "loss": 0.8884,
+      "step": 90
+    },
+    {
+      "epoch": 0.026807429693552164,
+      "grad_norm": 0.4857475757598877,
+      "learning_rate": 5.344827586206896e-06,
+      "loss": 0.497,
+      "step": 93
+    },
+    {
+      "epoch": 0.027672185490118364,
+      "grad_norm": 1.5475839376449585,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 0.669,
+      "step": 96
+    },
+    {
+      "epoch": 0.02853694128668456,
+      "grad_norm": 1.0966241359710693,
+      "learning_rate": 5.689655172413794e-06,
+      "loss": 0.5887,
+      "step": 99
+    },
+    {
+      "epoch": 0.028825193218873297,
+      "eval_loss": 0.6418702006340027,
+      "eval_mse": 0.6418701782226562,
+      "eval_runtime": 6.7842,
+      "eval_samples_per_second": 147.401,
+      "eval_steps_per_second": 18.425,
+      "step": 100
+    },
+    {
+      "epoch": 0.02940169708325076,
+      "grad_norm": 0.6968008875846863,
+      "learning_rate": 5.862068965517242e-06,
+      "loss": 0.7585,
+      "step": 102
+    },
+    {
+      "epoch": 0.03026645287981696,
+      "grad_norm": 1.3219984769821167,
+      "learning_rate": 6.03448275862069e-06,
+      "loss": 0.7782,
+      "step": 105
+    },
+    {
+      "epoch": 0.03113120867638316,
+      "grad_norm": 1.0195666551589966,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 0.6689,
+      "step": 108
+    },
+    {
+      "epoch": 0.031995964472949355,
+      "grad_norm": 1.4114892482757568,
+      "learning_rate": 6.379310344827587e-06,
+      "loss": 0.6057,
+      "step": 111
+    },
+    {
+      "epoch": 0.03286072026951556,
+      "grad_norm": 1.173322319984436,
+      "learning_rate": 6.551724137931035e-06,
+      "loss": 0.7234,
+      "step": 114
+    },
+    {
+      "epoch": 0.033725476066081755,
+      "grad_norm": 1.520411729812622,
+      "learning_rate": 6.724137931034484e-06,
+      "loss": 0.7145,
+      "step": 117
+    },
+    {
+      "epoch": 0.03459023186264795,
+      "grad_norm": 2.086540937423706,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.6265,
+      "step": 120
+    },
+    {
+      "epoch": 0.035454987659214156,
+      "grad_norm": 1.582171082496643,
+      "learning_rate": 7.0689655172413796e-06,
+      "loss": 0.6202,
+      "step": 123
+    },
+    {
+      "epoch": 0.03631974345578035,
+      "grad_norm": 1.1396665573120117,
+      "learning_rate": 7.241379310344828e-06,
+      "loss": 0.5784,
+      "step": 126
+    },
+    {
+      "epoch": 0.03718449925234655,
+      "grad_norm": 1.427815318107605,
+      "learning_rate": 7.413793103448277e-06,
+      "loss": 0.4815,
+      "step": 129
+    },
+    {
+      "epoch": 0.03804925504891275,
+      "grad_norm": 4.718282699584961,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 0.6151,
+      "step": 132
+    },
+    {
+      "epoch": 0.03891401084547895,
+      "grad_norm": 1.1084308624267578,
+      "learning_rate": 7.758620689655173e-06,
+      "loss": 0.548,
+      "step": 135
+    },
+    {
+      "epoch": 0.039778766642045146,
+      "grad_norm": 2.411567211151123,
+      "learning_rate": 7.93103448275862e-06,
+      "loss": 0.4664,
+      "step": 138
+    },
+    {
+      "epoch": 0.04064352243861134,
+      "grad_norm": 1.9377732276916504,
+      "learning_rate": 8.103448275862069e-06,
+      "loss": 0.4761,
+      "step": 141
+    },
+    {
+      "epoch": 0.04150827823517755,
+      "grad_norm": 1.6360706090927124,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 0.5101,
+      "step": 144
+    },
+    {
+      "epoch": 0.04237303403174374,
+      "grad_norm": 1.6306743621826172,
+      "learning_rate": 8.448275862068966e-06,
+      "loss": 0.4356,
+      "step": 147
+    },
+    {
+      "epoch": 0.04323778982830994,
+      "grad_norm": 2.038235664367676,
+      "learning_rate": 8.620689655172414e-06,
+      "loss": 0.4647,
+      "step": 150
+    },
+    {
+      "epoch": 0.044102545624876144,
+      "grad_norm": 2.968766450881958,
+      "learning_rate": 8.793103448275862e-06,
+      "loss": 0.5035,
+      "step": 153
+    },
+    {
+      "epoch": 0.04496730142144234,
+      "grad_norm": 1.407654047012329,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 0.4639,
+      "step": 156
+    },
+    {
+      "epoch": 0.04583205721800854,
+      "grad_norm": 2.141657590866089,
+      "learning_rate": 9.13793103448276e-06,
+      "loss": 0.4255,
+      "step": 159
+    },
+    {
+      "epoch": 0.04669681301457474,
+      "grad_norm": 2.2702648639678955,
+      "learning_rate": 9.310344827586207e-06,
+      "loss": 0.49,
+      "step": 162
+    },
+    {
+      "epoch": 0.04756156881114094,
+      "grad_norm": 1.1990652084350586,
+      "learning_rate": 9.482758620689655e-06,
+      "loss": 0.4145,
+      "step": 165
+    },
+    {
+      "epoch": 0.048426324607707134,
+      "grad_norm": 2.7780394554138184,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 0.4001,
+      "step": 168
+    },
+    {
+      "epoch": 0.04929108040427334,
+      "grad_norm": 1.6007318496704102,
+      "learning_rate": 9.827586206896553e-06,
+      "loss": 0.3719,
+      "step": 171
+    },
+    {
+      "epoch": 0.050155836200839535,
+      "grad_norm": 1.820497751235962,
+      "learning_rate": 1e-05,
+      "loss": 0.4183,
+      "step": 174
+    },
+    {
+      "epoch": 0.05102059199740573,
+      "grad_norm": 2.3770830631256104,
+      "learning_rate": 9.990895295902884e-06,
+      "loss": 0.4273,
+      "step": 177
+    },
+    {
+      "epoch": 0.05188534779397193,
+      "grad_norm": 1.9047311544418335,
+      "learning_rate": 9.981790591805767e-06,
+      "loss": 0.4622,
+      "step": 180
+    },
+    {
+      "epoch": 0.05275010359053813,
+      "grad_norm": 1.876373052597046,
+      "learning_rate": 9.972685887708651e-06,
+      "loss": 0.3132,
+      "step": 183
+    },
+    {
+      "epoch": 0.05361485938710433,
+      "grad_norm": 1.6469354629516602,
+      "learning_rate": 9.963581183611534e-06,
+      "loss": 0.3647,
+      "step": 186
+    },
+    {
+      "epoch": 0.054479615183670525,
+      "grad_norm": 1.3573085069656372,
+      "learning_rate": 9.954476479514417e-06,
+      "loss": 0.3267,
+      "step": 189
+    },
+    {
+      "epoch": 0.05534437098023673,
+      "grad_norm": 6.931013107299805,
+      "learning_rate": 9.9453717754173e-06,
+      "loss": 0.3981,
+      "step": 192
+    },
+    {
+      "epoch": 0.056209126776802926,
+      "grad_norm": 3.3258678913116455,
+      "learning_rate": 9.936267071320182e-06,
+      "loss": 0.3855,
+      "step": 195
+    },
+    {
+      "epoch": 0.05707388257336912,
+      "grad_norm": 4.0184550285339355,
+      "learning_rate": 9.927162367223067e-06,
+      "loss": 0.371,
+      "step": 198
+    },
+    {
+      "epoch": 0.057650386437746594,
+      "eval_loss": 0.34387460350990295,
+      "eval_mse": 0.3438746197223663,
+      "eval_runtime": 6.7538,
+      "eval_samples_per_second": 148.065,
+      "eval_steps_per_second": 18.508,
+      "step": 200
+    },
+    {
+      "epoch": 0.057938638369935326,
+      "grad_norm": 1.7570185661315918,
+      "learning_rate": 9.91805766312595e-06,
+      "loss": 0.4019,
+      "step": 201
+    },
+    {
+      "epoch": 0.05880339416650152,
+      "grad_norm": 3.925257444381714,
+      "learning_rate": 9.908952959028833e-06,
+      "loss": 0.3352,
+      "step": 204
+    },
+    {
+      "epoch": 0.05966814996306772,
+      "grad_norm": 1.5225563049316406,
+      "learning_rate": 9.899848254931715e-06,
+      "loss": 0.3081,
+      "step": 207
+    },
+    {
+      "epoch": 0.06053290575963392,
+      "grad_norm": 1.902543067932129,
+      "learning_rate": 9.890743550834598e-06,
+      "loss": 0.3908,
+      "step": 210
+    },
+    {
+      "epoch": 0.06139766155620012,
+      "grad_norm": 1.2444404363632202,
+      "learning_rate": 9.881638846737481e-06,
+      "loss": 0.31,
+      "step": 213
+    },
+    {
+      "epoch": 0.06226241735276632,
+      "grad_norm": 1.8852041959762573,
+      "learning_rate": 9.872534142640366e-06,
+      "loss": 0.297,
+      "step": 216
+    },
+    {
+      "epoch": 0.06312717314933251,
+      "grad_norm": 5.955360412597656,
+      "learning_rate": 9.863429438543249e-06,
+      "loss": 0.3639,
+      "step": 219
+    },
+    {
+      "epoch": 0.06399192894589871,
+      "grad_norm": 1.8612481355667114,
+      "learning_rate": 9.854324734446131e-06,
+      "loss": 0.3808,
+      "step": 222
+    },
+    {
+      "epoch": 0.06485668474246492,
+      "grad_norm": 1.4975637197494507,
+      "learning_rate": 9.845220030349014e-06,
+      "loss": 0.3694,
+      "step": 225
+    },
+    {
+      "epoch": 0.06572144053903112,
+      "grad_norm": 1.5896477699279785,
+      "learning_rate": 9.836115326251897e-06,
+      "loss": 0.3056,
+      "step": 228
+    },
+    {
+      "epoch": 0.06658619633559731,
+      "grad_norm": 3.6155972480773926,
+      "learning_rate": 9.827010622154782e-06,
+      "loss": 0.3084,
+      "step": 231
+    },
+    {
+      "epoch": 0.06745095213216351,
+      "grad_norm": 4.709763526916504,
+      "learning_rate": 9.817905918057664e-06,
+      "loss": 0.3949,
+      "step": 234
+    },
+    {
+      "epoch": 0.06831570792872971,
+      "grad_norm": 3.7941088676452637,
+      "learning_rate": 9.808801213960547e-06,
+      "loss": 0.3213,
+      "step": 237
+    },
+    {
+      "epoch": 0.0691804637252959,
+      "grad_norm": 2.548584461212158,
+      "learning_rate": 9.79969650986343e-06,
+      "loss": 0.3335,
+      "step": 240
+    },
+    {
+      "epoch": 0.0700452195218621,
+      "grad_norm": 1.5992470979690552,
+      "learning_rate": 9.790591805766313e-06,
+      "loss": 0.2688,
+      "step": 243
+    },
+    {
+      "epoch": 0.07090997531842831,
+      "grad_norm": 2.99310564994812,
+      "learning_rate": 9.781487101669198e-06,
+      "loss": 0.3488,
+      "step": 246
+    },
+    {
+      "epoch": 0.07177473111499451,
+      "grad_norm": 2.987074613571167,
+      "learning_rate": 9.77238239757208e-06,
+      "loss": 0.2703,
+      "step": 249
+    },
+    {
+      "epoch": 0.0726394869115607,
+      "grad_norm": 5.408164978027344,
+      "learning_rate": 9.763277693474963e-06,
+      "loss": 0.2854,
+      "step": 252
+    },
+    {
+      "epoch": 0.0735042427081269,
+      "grad_norm": 1.6425719261169434,
+      "learning_rate": 9.754172989377846e-06,
+      "loss": 0.2824,
+      "step": 255
+    },
+    {
+      "epoch": 0.0743689985046931,
+      "grad_norm": 1.8134018182754517,
+      "learning_rate": 9.745068285280729e-06,
+      "loss": 0.3286,
+      "step": 258
+    },
+    {
+      "epoch": 0.0752337543012593,
+      "grad_norm": 5.619387626647949,
+      "learning_rate": 9.735963581183613e-06,
+      "loss": 0.3226,
+      "step": 261
+    },
+    {
+      "epoch": 0.0760985100978255,
+      "grad_norm": 2.6264755725860596,
+      "learning_rate": 9.726858877086496e-06,
+      "loss": 0.3442,
+      "step": 264
+    },
+    {
+      "epoch": 0.0769632658943917,
+      "grad_norm": 3.537142515182495,
+      "learning_rate": 9.717754172989379e-06,
+      "loss": 0.2687,
+      "step": 267
+    },
+    {
+      "epoch": 0.0778280216909579,
+      "grad_norm": 2.1801705360412598,
+      "learning_rate": 9.708649468892262e-06,
+      "loss": 0.3705,
+      "step": 270
+    },
+    {
+      "epoch": 0.0786927774875241,
+      "grad_norm": 6.869683742523193,
+      "learning_rate": 9.699544764795145e-06,
+      "loss": 0.3777,
+      "step": 273
+    },
+    {
+      "epoch": 0.07955753328409029,
+      "grad_norm": 3.4050021171569824,
+      "learning_rate": 9.690440060698028e-06,
+      "loss": 0.3265,
+      "step": 276
+    },
+    {
+      "epoch": 0.08042228908065649,
+      "grad_norm": 2.8107333183288574,
+      "learning_rate": 9.681335356600912e-06,
+      "loss": 0.3286,
+      "step": 279
+    },
+    {
+      "epoch": 0.08128704487722269,
+      "grad_norm": 1.5242034196853638,
+      "learning_rate": 9.672230652503795e-06,
+      "loss": 0.2786,
+      "step": 282
+    },
+    {
+      "epoch": 0.0821518006737889,
+      "grad_norm": 22.8824462890625,
+      "learning_rate": 9.663125948406678e-06,
+      "loss": 0.3706,
+      "step": 285
+    },
+    {
+      "epoch": 0.0830165564703551,
+      "grad_norm": 4.904447555541992,
+      "learning_rate": 9.65402124430956e-06,
+      "loss": 0.298,
+      "step": 288
+    },
+    {
+      "epoch": 0.08388131226692129,
+      "grad_norm": 4.405270576477051,
+      "learning_rate": 9.644916540212444e-06,
+      "loss": 0.3625,
+      "step": 291
+    },
+    {
+      "epoch": 0.08474606806348749,
+      "grad_norm": 6.424873352050781,
+      "learning_rate": 9.635811836115328e-06,
+      "loss": 0.3246,
+      "step": 294
+    },
+    {
+      "epoch": 0.08561082386005368,
+      "grad_norm": 1.9719147682189941,
+      "learning_rate": 9.626707132018211e-06,
+      "loss": 0.3014,
+      "step": 297
+    },
+    {
+      "epoch": 0.08647557965661988,
+      "grad_norm": 5.9254021644592285,
+      "learning_rate": 9.617602427921094e-06,
+      "loss": 0.3607,
+      "step": 300
+    },
+    {
+      "epoch": 0.08647557965661988,
+      "eval_loss": 0.28438636660575867,
+      "eval_mse": 0.2843863361030817,
+      "eval_runtime": 6.6145,
+      "eval_samples_per_second": 151.183,
+      "eval_steps_per_second": 18.898,
+      "step": 300
+    },
+    {
+      "epoch": 0.08734033545318609,
+      "grad_norm": 3.4420008659362793,
+      "learning_rate": 9.608497723823977e-06,
+      "loss": 0.2631,
+      "step": 303
+    },
+    {
+      "epoch": 0.08820509124975229,
+      "grad_norm": 2.4832332134246826,
+      "learning_rate": 9.59939301972686e-06,
+      "loss": 0.2985,
+      "step": 306
+    },
+    {
+      "epoch": 0.08906984704631848,
+      "grad_norm": 3.3724935054779053,
+      "learning_rate": 9.590288315629744e-06,
+      "loss": 0.2703,
+      "step": 309
+    },
+    {
+      "epoch": 0.08993460284288468,
+      "grad_norm": 8.569412231445312,
+      "learning_rate": 9.581183611532627e-06,
+      "loss": 0.3068,
+      "step": 312
+    },
+    {
+      "epoch": 0.09079935863945088,
+      "grad_norm": 1.9817373752593994,
+      "learning_rate": 9.57207890743551e-06,
+      "loss": 0.3271,
+      "step": 315
+    },
+    {
+      "epoch": 0.09166411443601707,
+      "grad_norm": 2.6405210494995117,
+      "learning_rate": 9.562974203338393e-06,
+      "loss": 0.2454,
+      "step": 318
+    },
+    {
+      "epoch": 0.09252887023258327,
+      "grad_norm": 1.6005452871322632,
+      "learning_rate": 9.553869499241275e-06,
+      "loss": 0.2506,
+      "step": 321
+    },
+    {
+      "epoch": 0.09339362602914948,
+      "grad_norm": 1.67816162109375,
+      "learning_rate": 9.54476479514416e-06,
+      "loss": 0.2503,
+      "step": 324
+    },
+    {
+      "epoch": 0.09425838182571568,
+      "grad_norm": 1.4401655197143555,
+      "learning_rate": 9.535660091047043e-06,
+      "loss": 0.2322,
+      "step": 327
+    },
+    {
+      "epoch": 0.09512313762228188,
+      "grad_norm": 2.919785261154175,
+      "learning_rate": 9.526555386949926e-06,
+      "loss": 0.2746,
+      "step": 330
+    },
+    {
+      "epoch": 0.09598789341884807,
+      "grad_norm": 4.454317569732666,
+      "learning_rate": 9.517450682852808e-06,
+      "loss": 0.2907,
+      "step": 333
+    },
+    {
+      "epoch": 0.09685264921541427,
+      "grad_norm": 4.585294246673584,
+      "learning_rate": 9.508345978755691e-06,
+      "loss": 0.2873,
+      "step": 336
+    },
+    {
+      "epoch": 0.09771740501198047,
+      "grad_norm": 2.247422218322754,
+      "learning_rate": 9.499241274658574e-06,
+      "loss": 0.2798,
+      "step": 339
+    },
+    {
+      "epoch": 0.09858216080854668,
+      "grad_norm": 3.6044836044311523,
+      "learning_rate": 9.490136570561459e-06,
+      "loss": 0.3284,
+      "step": 342
+    },
+    {
+      "epoch": 0.09944691660511287,
+      "grad_norm": 4.609151363372803,
+      "learning_rate": 9.481031866464341e-06,
+      "loss": 0.2274,
+      "step": 345
+    },
+    {
+      "epoch": 0.10031167240167907,
+      "grad_norm": 4.793229103088379,
+      "learning_rate": 9.471927162367224e-06,
+      "loss": 0.2734,
+      "step": 348
+    },
+    {
+      "epoch": 0.10117642819824527,
+      "grad_norm": 4.507264614105225,
+      "learning_rate": 9.462822458270107e-06,
+      "loss": 0.3082,
+      "step": 351
+    },
+    {
+      "epoch": 0.10204118399481146,
+      "grad_norm": 6.129451274871826,
+      "learning_rate": 9.45371775417299e-06,
+      "loss": 0.2617,
+      "step": 354
+    },
+    {
+      "epoch": 0.10290593979137766,
+      "grad_norm": 2.631593704223633,
+      "learning_rate": 9.444613050075875e-06,
+      "loss": 0.2565,
+      "step": 357
+    },
+    {
+      "epoch": 0.10377069558794386,
+      "grad_norm": 4.379823684692383,
+      "learning_rate": 9.435508345978757e-06,
+      "loss": 0.3235,
+      "step": 360
+    },
+    {
+      "epoch": 0.10463545138451007,
+      "grad_norm": 2.077354907989502,
+      "learning_rate": 9.42640364188164e-06,
+      "loss": 0.2938,
+      "step": 363
+    },
+    {
+      "epoch": 0.10550020718107626,
+      "grad_norm": 5.108116626739502,
+      "learning_rate": 9.417298937784523e-06,
+      "loss": 0.2698,
+      "step": 366
+    },
+    {
+      "epoch": 0.10636496297764246,
+      "grad_norm": 7.464448928833008,
+      "learning_rate": 9.408194233687406e-06,
+      "loss": 0.2703,
+      "step": 369
+    },
+    {
+      "epoch": 0.10722971877420866,
+      "grad_norm": 1.855411410331726,
+      "learning_rate": 9.399089529590289e-06,
+      "loss": 0.2958,
+      "step": 372
+    },
+    {
+      "epoch": 0.10809447457077485,
+      "grad_norm": 5.284719944000244,
+      "learning_rate": 9.389984825493173e-06,
+      "loss": 0.2703,
+      "step": 375
+    },
+    {
+      "epoch": 0.10895923036734105,
+      "grad_norm": 2.494473457336426,
+      "learning_rate": 9.380880121396056e-06,
+      "loss": 0.2698,
+      "step": 378
+    },
+    {
+      "epoch": 0.10982398616390726,
+      "grad_norm": 2.0765345096588135,
+      "learning_rate": 9.371775417298939e-06,
+      "loss": 0.2726,
+      "step": 381
+    },
+    {
+      "epoch": 0.11068874196047346,
+      "grad_norm": 1.893574595451355,
+      "learning_rate": 9.362670713201822e-06,
+      "loss": 0.2739,
+      "step": 384
+    },
+    {
+      "epoch": 0.11155349775703965,
+      "grad_norm": 2.0016255378723145,
+      "learning_rate": 9.353566009104705e-06,
+      "loss": 0.2624,
+      "step": 387
+    },
+    {
+      "epoch": 0.11241825355360585,
+      "grad_norm": 2.99924898147583,
+      "learning_rate": 9.344461305007587e-06,
+      "loss": 0.3098,
+      "step": 390
+    },
+    {
+      "epoch": 0.11328300935017205,
+      "grad_norm": 1.666891098022461,
+      "learning_rate": 9.335356600910472e-06,
+      "loss": 0.2451,
+      "step": 393
+    },
+    {
+      "epoch": 0.11414776514673824,
+      "grad_norm": 2.8993024826049805,
+      "learning_rate": 9.326251896813355e-06,
+      "loss": 0.271,
+      "step": 396
+    },
+    {
+      "epoch": 0.11501252094330444,
+      "grad_norm": 5.040359973907471,
+      "learning_rate": 9.317147192716238e-06,
+      "loss": 0.2576,
+      "step": 399
+    },
+    {
+      "epoch": 0.11530077287549319,
+      "eval_loss": 0.2588765025138855,
+      "eval_mse": 0.25887649209996744,
+      "eval_runtime": 6.6586,
+      "eval_samples_per_second": 150.181,
+      "eval_steps_per_second": 18.773,
+      "step": 400
+    },
+    {
+      "epoch": 0.11587727673987065,
+      "grad_norm": 3.234560251235962,
+      "learning_rate": 9.30804248861912e-06,
+      "loss": 0.2702,
+      "step": 402
+    },
+    {
+      "epoch": 0.11674203253643685,
+      "grad_norm": 2.8497729301452637,
+      "learning_rate": 9.298937784522003e-06,
+      "loss": 0.2371,
+      "step": 405
+    },
+    {
+      "epoch": 0.11760678833300305,
+      "grad_norm": 1.3214294910430908,
+      "learning_rate": 9.289833080424886e-06,
+      "loss": 0.2262,
+      "step": 408
+    },
+    {
+      "epoch": 0.11847154412956924,
+      "grad_norm": 3.5736958980560303,
+      "learning_rate": 9.28072837632777e-06,
+      "loss": 0.2619,
+      "step": 411
+    },
+    {
+      "epoch": 0.11933629992613544,
+      "grad_norm": 3.478178024291992,
+      "learning_rate": 9.271623672230654e-06,
+      "loss": 0.3061,
+      "step": 414
+    },
+    {
+      "epoch": 0.12020105572270164,
+      "grad_norm": 2.523387908935547,
+      "learning_rate": 9.262518968133536e-06,
+      "loss": 0.3232,
+      "step": 417
+    },
+    {
+      "epoch": 0.12106581151926785,
+      "grad_norm": 2.1046786308288574,
+      "learning_rate": 9.25341426403642e-06,
+      "loss": 0.2815,
+      "step": 420
+    },
+    {
+      "epoch": 0.12193056731583404,
+      "grad_norm": 4.513411045074463,
+      "learning_rate": 9.244309559939302e-06,
+      "loss": 0.3053,
+      "step": 423
+    },
+    {
+      "epoch": 0.12279532311240024,
+      "grad_norm": 2.628030776977539,
+      "learning_rate": 9.235204855842187e-06,
+      "loss": 0.2714,
+      "step": 426
+    },
+    {
+      "epoch": 0.12366007890896644,
+      "grad_norm": 6.008927345275879,
+      "learning_rate": 9.22610015174507e-06,
+      "loss": 0.3243,
+      "step": 429
+    },
+    {
+      "epoch": 0.12452483470553263,
+      "grad_norm": 6.236274242401123,
+      "learning_rate": 9.216995447647952e-06,
+      "loss": 0.2916,
+      "step": 432
+    },
+    {
+      "epoch": 0.12538959050209883,
+      "grad_norm": 2.30412220954895,
+      "learning_rate": 9.207890743550835e-06,
+      "loss": 0.2731,
+      "step": 435
+    },
+    {
+      "epoch": 0.12625434629866503,
+      "grad_norm": 3.3161492347717285,
+      "learning_rate": 9.198786039453718e-06,
+      "loss": 0.2508,
+      "step": 438
+    },
+    {
+      "epoch": 0.12711910209523122,
+      "grad_norm": 4.023074626922607,
+      "learning_rate": 9.189681335356601e-06,
+      "loss": 0.2973,
+      "step": 441
+    },
+    {
+      "epoch": 0.12798385789179742,
+      "grad_norm": 2.486236333847046,
+      "learning_rate": 9.180576631259485e-06,
+      "loss": 0.235,
+      "step": 444
+    },
+    {
+      "epoch": 0.12884861368836362,
+      "grad_norm": 4.445496082305908,
+      "learning_rate": 9.171471927162368e-06,
+      "loss": 0.2519,
+      "step": 447
+    },
+    {
+      "epoch": 0.12971336948492984,
+      "grad_norm": 1.629809021949768,
+      "learning_rate": 9.162367223065251e-06,
+      "loss": 0.2664,
+      "step": 450
+    },
+    {
+      "epoch": 0.13057812528149604,
+      "grad_norm": 3.0351674556732178,
+      "learning_rate": 9.153262518968134e-06,
+      "loss": 0.2433,
+      "step": 453
+    },
+    {
+      "epoch": 0.13144288107806223,
+      "grad_norm": 1.9163283109664917,
+      "learning_rate": 9.144157814871017e-06,
+      "loss": 0.2997,
+      "step": 456
+    },
+    {
+      "epoch": 0.13230763687462843,
+      "grad_norm": 3.978429079055786,
+      "learning_rate": 9.1350531107739e-06,
+      "loss": 0.2969,
+      "step": 459
+    },
+    {
+      "epoch": 0.13317239267119463,
+      "grad_norm": 1.7428011894226074,
+      "learning_rate": 9.125948406676784e-06,
+      "loss": 0.2721,
+      "step": 462
+    },
+    {
+      "epoch": 0.13403714846776082,
+      "grad_norm": 1.462106704711914,
+      "learning_rate": 9.116843702579667e-06,
+      "loss": 0.3179,
+      "step": 465
+    },
+    {
+      "epoch": 0.13490190426432702,
+      "grad_norm": 4.226785659790039,
+      "learning_rate": 9.10773899848255e-06,
+      "loss": 0.262,
+      "step": 468
+    },
+    {
+      "epoch": 0.13576666006089322,
+      "grad_norm": 3.227842330932617,
+      "learning_rate": 9.098634294385433e-06,
+      "loss": 0.2602,
+      "step": 471
+    },
+    {
+      "epoch": 0.13663141585745942,
+      "grad_norm": 4.708644866943359,
+      "learning_rate": 9.089529590288316e-06,
+      "loss": 0.2502,
+      "step": 474
+    },
+    {
+      "epoch": 0.1374961716540256,
+      "grad_norm": 3.478773832321167,
+      "learning_rate": 9.080424886191198e-06,
+      "loss": 0.2799,
+      "step": 477
+    },
+    {
+      "epoch": 0.1383609274505918,
+      "grad_norm": 5.991547107696533,
+      "learning_rate": 9.071320182094083e-06,
+      "loss": 0.3103,
+      "step": 480
+    },
+    {
+      "epoch": 0.139225683247158,
+      "grad_norm": 5.446369647979736,
+      "learning_rate": 9.062215477996966e-06,
+      "loss": 0.2834,
+      "step": 483
+    },
+    {
+      "epoch": 0.1400904390437242,
+      "grad_norm": 4.723580360412598,
+      "learning_rate": 9.053110773899849e-06,
+      "loss": 0.244,
+      "step": 486
+    },
+    {
+      "epoch": 0.14095519484029043,
+      "grad_norm": 2.518148183822632,
+      "learning_rate": 9.044006069802731e-06,
+      "loss": 0.3102,
+      "step": 489
+    },
+    {
+      "epoch": 0.14181995063685662,
+      "grad_norm": 4.1146039962768555,
+      "learning_rate": 9.034901365705614e-06,
+      "loss": 0.2913,
+      "step": 492
+    },
+    {
+      "epoch": 0.14268470643342282,
+      "grad_norm": 2.6333212852478027,
+      "learning_rate": 9.025796661608497e-06,
+      "loss": 0.2195,
+      "step": 495
+    },
+    {
+      "epoch": 0.14354946222998902,
+      "grad_norm": 3.344228506088257,
+      "learning_rate": 9.016691957511382e-06,
+      "loss": 0.2822,
+      "step": 498
+    },
+    {
+      "epoch": 0.14412596609436648,
+      "eval_loss": 0.2707275450229645,
+      "eval_mse": 0.27072753977278263,
+      "eval_runtime": 6.638,
+      "eval_samples_per_second": 150.647,
+      "eval_steps_per_second": 18.831,
+      "step": 500
+    },
+    {
+      "epoch": 0.1444142180265552,
+      "grad_norm": 2.839517593383789,
+      "learning_rate": 9.007587253414265e-06,
+      "loss": 0.2469,
+      "step": 501
+    },
+    {
+      "epoch": 0.1452789738231214,
+      "grad_norm": 3.71785569190979,
+      "learning_rate": 8.998482549317147e-06,
+      "loss": 0.2695,
+      "step": 504
+    },
+    {
+      "epoch": 0.1461437296196876,
+      "grad_norm": 2.9626924991607666,
+      "learning_rate": 8.98937784522003e-06,
+      "loss": 0.233,
+      "step": 507
+    },
+    {
+      "epoch": 0.1470084854162538,
+      "grad_norm": 2.0274481773376465,
+      "learning_rate": 8.980273141122913e-06,
+      "loss": 0.2854,
+      "step": 510
+    },
+    {
+      "epoch": 0.14787324121282,
+      "grad_norm": 2.5144805908203125,
+      "learning_rate": 8.971168437025798e-06,
+      "loss": 0.2545,
+      "step": 513
+    },
+    {
+      "epoch": 0.1487379970093862,
+      "grad_norm": 3.106039047241211,
+      "learning_rate": 8.96206373292868e-06,
+      "loss": 0.2305,
+      "step": 516
+    },
+    {
+      "epoch": 0.1496027528059524,
+      "grad_norm": 3.3986117839813232,
+      "learning_rate": 8.952959028831563e-06,
+      "loss": 0.2564,
+      "step": 519
+    },
+    {
+      "epoch": 0.1504675086025186,
+      "grad_norm": 2.877206325531006,
+      "learning_rate": 8.943854324734446e-06,
+      "loss": 0.2781,
+      "step": 522
+    },
+    {
+      "epoch": 0.1513322643990848,
+      "grad_norm": 2.561119556427002,
+      "learning_rate": 8.934749620637329e-06,
+      "loss": 0.2666,
+      "step": 525
+    },
+    {
+      "epoch": 0.152197020195651,
+      "grad_norm": 2.566633939743042,
+      "learning_rate": 8.925644916540213e-06,
+      "loss": 0.2791,
+      "step": 528
+    },
+    {
+      "epoch": 0.1530617759922172,
+      "grad_norm": 2.1075491905212402,
+      "learning_rate": 8.916540212443096e-06,
+      "loss": 0.2276,
+      "step": 531
+    },
+    {
+      "epoch": 0.1539265317887834,
+      "grad_norm": 3.239712715148926,
+      "learning_rate": 8.90743550834598e-06,
+      "loss": 0.2749,
+      "step": 534
+    },
+    {
+      "epoch": 0.1547912875853496,
+      "grad_norm": 6.005987167358398,
+      "learning_rate": 8.898330804248862e-06,
+      "loss": 0.2395,
+      "step": 537
+    },
+    {
+      "epoch": 0.1556560433819158,
+      "grad_norm": 1.6621935367584229,
+      "learning_rate": 8.889226100151745e-06,
+      "loss": 0.2616,
+      "step": 540
+    },
+    {
+      "epoch": 0.156520799178482,
+      "grad_norm": 3.1361093521118164,
+      "learning_rate": 8.880121396054628e-06,
+      "loss": 0.2712,
+      "step": 543
+    },
+    {
+      "epoch": 0.1573855549750482,
+      "grad_norm": 1.8823013305664062,
+      "learning_rate": 8.871016691957512e-06,
+      "loss": 0.2571,
+      "step": 546
+    },
+    {
+      "epoch": 0.1582503107716144,
+      "grad_norm": 3.6857988834381104,
+      "learning_rate": 8.861911987860395e-06,
+      "loss": 0.2552,
+      "step": 549
+    },
+    {
+      "epoch": 0.15911506656818059,
+      "grad_norm": 2.26597261428833,
+      "learning_rate": 8.852807283763278e-06,
+      "loss": 0.2239,
+      "step": 552
+    },
+    {
+      "epoch": 0.15997982236474678,
+      "grad_norm": 5.791572570800781,
+      "learning_rate": 8.84370257966616e-06,
+      "loss": 0.2859,
+      "step": 555
+    },
+    {
+      "epoch": 0.16084457816131298,
+      "grad_norm": 3.4344265460968018,
+      "learning_rate": 8.834597875569044e-06,
+      "loss": 0.3055,
+      "step": 558
+    },
+    {
+      "epoch": 0.16170933395787918,
+      "grad_norm": 3.98288631439209,
+      "learning_rate": 8.825493171471928e-06,
+      "loss": 0.2698,
+      "step": 561
+    },
+    {
+      "epoch": 0.16257408975444537,
+      "grad_norm": 7.429836273193359,
+      "learning_rate": 8.816388467374811e-06,
+      "loss": 0.3207,
+      "step": 564
+    },
+    {
+      "epoch": 0.1634388455510116,
+      "grad_norm": 4.021480560302734,
+      "learning_rate": 8.807283763277694e-06,
+      "loss": 0.2477,
+      "step": 567
+    },
+    {
+      "epoch": 0.1643036013475778,
+      "grad_norm": 2.619497537612915,
+      "learning_rate": 8.798179059180577e-06,
+      "loss": 0.2152,
+      "step": 570
+    },
+    {
+      "epoch": 0.165168357144144,
+      "grad_norm": 2.073925018310547,
+      "learning_rate": 8.78907435508346e-06,
+      "loss": 0.2557,
+      "step": 573
+    },
+    {
+      "epoch": 0.1660331129407102,
+      "grad_norm": 2.250293493270874,
+      "learning_rate": 8.779969650986344e-06,
+      "loss": 0.2294,
+      "step": 576
+    },
+    {
+      "epoch": 0.16689786873727638,
+      "grad_norm": 2.9747421741485596,
+      "learning_rate": 8.770864946889227e-06,
+      "loss": 0.2461,
+      "step": 579
+    },
+    {
+      "epoch": 0.16776262453384258,
+      "grad_norm": 1.8815991878509521,
+      "learning_rate": 8.76176024279211e-06,
+      "loss": 0.2698,
+      "step": 582
+    },
+    {
+      "epoch": 0.16862738033040878,
+      "grad_norm": 2.1905224323272705,
+      "learning_rate": 8.752655538694993e-06,
+      "loss": 0.2631,
+      "step": 585
+    },
+    {
+      "epoch": 0.16949213612697497,
+      "grad_norm": 3.6808903217315674,
+      "learning_rate": 8.743550834597875e-06,
+      "loss": 0.2341,
+      "step": 588
+    },
+    {
+      "epoch": 0.17035689192354117,
+      "grad_norm": 2.905963897705078,
+      "learning_rate": 8.73444613050076e-06,
+      "loss": 0.2295,
+      "step": 591
+    },
+    {
+      "epoch": 0.17122164772010737,
+      "grad_norm": 5.486540794372559,
+      "learning_rate": 8.725341426403643e-06,
+      "loss": 0.2816,
+      "step": 594
+    },
+    {
+      "epoch": 0.17208640351667356,
+      "grad_norm": 1.8525919914245605,
+      "learning_rate": 8.716236722306526e-06,
+      "loss": 0.2197,
+      "step": 597
+    },
+    {
+      "epoch": 0.17295115931323976,
+      "grad_norm": 5.908591270446777,
+      "learning_rate": 8.707132018209408e-06,
+      "loss": 0.2908,
+      "step": 600
+    },
+    {
+      "epoch": 0.17295115931323976,
+      "eval_loss": 0.23815499246120453,
+      "eval_mse": 0.23815500601008535,
+      "eval_runtime": 6.6897,
+      "eval_samples_per_second": 149.484,
+      "eval_steps_per_second": 18.685,
+      "step": 600
+    },
+    {
+      "epoch": 0.17381591510980596,
+      "grad_norm": 3.1334948539733887,
+      "learning_rate": 8.698027314112291e-06,
+      "loss": 0.253,
+      "step": 603
+    },
+    {
+      "epoch": 0.17468067090637218,
+      "grad_norm": 2.314412832260132,
+      "learning_rate": 8.688922610015174e-06,
+      "loss": 0.2631,
+      "step": 606
+    },
+    {
+      "epoch": 0.17554542670293838,
+      "grad_norm": 3.483959436416626,
+      "learning_rate": 8.679817905918059e-06,
+      "loss": 0.2452,
+      "step": 609
+    },
+    {
+      "epoch": 0.17641018249950458,
+      "grad_norm": 2.2747623920440674,
+      "learning_rate": 8.670713201820942e-06,
+      "loss": 0.2462,
+      "step": 612
+    },
+    {
+      "epoch": 0.17727493829607077,
+      "grad_norm": 5.517392635345459,
+      "learning_rate": 8.661608497723824e-06,
+      "loss": 0.2707,
+      "step": 615
+    },
+    {
+      "epoch": 0.17813969409263697,
+      "grad_norm": 2.7062125205993652,
+      "learning_rate": 8.652503793626707e-06,
+      "loss": 0.2324,
+      "step": 618
+    },
+    {
+      "epoch": 0.17900444988920317,
+      "grad_norm": 2.712933301925659,
+      "learning_rate": 8.64339908952959e-06,
+      "loss": 0.2914,
+      "step": 621
+    },
+    {
+      "epoch": 0.17986920568576936,
+      "grad_norm": 3.0349957942962646,
+      "learning_rate": 8.634294385432475e-06,
+      "loss": 0.245,
+      "step": 624
+    },
+    {
+      "epoch": 0.18073396148233556,
+      "grad_norm": 1.533530592918396,
+      "learning_rate": 8.625189681335357e-06,
+      "loss": 0.245,
+      "step": 627
+    },
+    {
+      "epoch": 0.18159871727890176,
+      "grad_norm": 1.5070098638534546,
+      "learning_rate": 8.61608497723824e-06,
+      "loss": 0.252,
+      "step": 630
+    },
+    {
+      "epoch": 0.18246347307546795,
+      "grad_norm": 1.803514003753662,
+      "learning_rate": 8.606980273141123e-06,
+      "loss": 0.2395,
+      "step": 633
+    },
+    {
+      "epoch": 0.18332822887203415,
+      "grad_norm": 1.8047112226486206,
+      "learning_rate": 8.597875569044006e-06,
+      "loss": 0.2731,
+      "step": 636
+    },
+    {
+      "epoch": 0.18419298466860035,
+      "grad_norm": 3.090371608734131,
+      "learning_rate": 8.58877086494689e-06,
+      "loss": 0.2314,
+      "step": 639
+    },
+    {
+      "epoch": 0.18505774046516654,
+      "grad_norm": 2.2020678520202637,
+      "learning_rate": 8.579666160849773e-06,
+      "loss": 0.2828,
+      "step": 642
+    },
+    {
+      "epoch": 0.18592249626173277,
+      "grad_norm": 1.8086636066436768,
+      "learning_rate": 8.570561456752656e-06,
+      "loss": 0.2736,
+      "step": 645
+    },
+    {
+      "epoch": 0.18678725205829896,
+      "grad_norm": 1.9154764413833618,
+      "learning_rate": 8.561456752655539e-06,
+      "loss": 0.2422,
+      "step": 648
+    },
+    {
+      "epoch": 0.18765200785486516,
+      "grad_norm": 2.6091620922088623,
+      "learning_rate": 8.552352048558422e-06,
+      "loss": 0.278,
+      "step": 651
+    },
+    {
+      "epoch": 0.18851676365143136,
+      "grad_norm": 4.057301998138428,
+      "learning_rate": 8.543247344461306e-06,
+      "loss": 0.2268,
+      "step": 654
+    },
+    {
+      "epoch": 0.18938151944799755,
+      "grad_norm": 1.6707180738449097,
+      "learning_rate": 8.53414264036419e-06,
+      "loss": 0.2657,
+      "step": 657
+    },
+    {
+      "epoch": 0.19024627524456375,
+      "grad_norm": 4.327409744262695,
+      "learning_rate": 8.525037936267072e-06,
+      "loss": 0.2235,
+      "step": 660
+    },
+    {
+      "epoch": 0.19111103104112995,
+      "grad_norm": 3.991241931915283,
+      "learning_rate": 8.515933232169955e-06,
+      "loss": 0.2429,
+      "step": 663
+    },
+    {
+      "epoch": 0.19197578683769614,
+      "grad_norm": 1.742564082145691,
+      "learning_rate": 8.506828528072838e-06,
+      "loss": 0.2245,
+      "step": 666
+    },
+    {
+      "epoch": 0.19284054263426234,
+      "grad_norm": 2.362626791000366,
+      "learning_rate": 8.49772382397572e-06,
+      "loss": 0.2689,
+      "step": 669
+    },
+    {
+      "epoch": 0.19370529843082854,
+      "grad_norm": 1.4896934032440186,
+      "learning_rate": 8.488619119878605e-06,
+      "loss": 0.2669,
+      "step": 672
+    },
+    {
+      "epoch": 0.19457005422739473,
+      "grad_norm": 1.8312315940856934,
+      "learning_rate": 8.479514415781488e-06,
+      "loss": 0.2255,
+      "step": 675
+    },
+    {
+      "epoch": 0.19543481002396093,
+      "grad_norm": 3.9799551963806152,
+      "learning_rate": 8.470409711684371e-06,
+      "loss": 0.2152,
+      "step": 678
+    },
+    {
+      "epoch": 0.19629956582052713,
+      "grad_norm": 1.9584782123565674,
+      "learning_rate": 8.461305007587254e-06,
+      "loss": 0.2721,
+      "step": 681
+    },
+    {
+      "epoch": 0.19716432161709335,
+      "grad_norm": 3.361952781677246,
+      "learning_rate": 8.452200303490137e-06,
+      "loss": 0.265,
+      "step": 684
+    },
+    {
+      "epoch": 0.19802907741365955,
+      "grad_norm": 2.125466823577881,
+      "learning_rate": 8.443095599393021e-06,
+      "loss": 0.2473,
+      "step": 687
+    },
+    {
+      "epoch": 0.19889383321022575,
+      "grad_norm": 4.599373817443848,
+      "learning_rate": 8.433990895295904e-06,
+      "loss": 0.2201,
+      "step": 690
+    },
+    {
+      "epoch": 0.19975858900679194,
+      "grad_norm": 2.139647960662842,
+      "learning_rate": 8.424886191198787e-06,
+      "loss": 0.2647,
+      "step": 693
+    },
+    {
+      "epoch": 0.20062334480335814,
+      "grad_norm": 1.7437653541564941,
+      "learning_rate": 8.41578148710167e-06,
+      "loss": 0.2242,
+      "step": 696
+    },
+    {
+      "epoch": 0.20148810059992434,
+      "grad_norm": 4.154083728790283,
+      "learning_rate": 8.406676783004552e-06,
+      "loss": 0.2258,
+      "step": 699
+    },
+    {
+      "epoch": 0.20177635253211307,
+      "eval_loss": 0.2404525727033615,
+      "eval_mse": 0.24045259381830691,
+      "eval_runtime": 6.5158,
+      "eval_samples_per_second": 153.474,
+      "eval_steps_per_second": 19.184,
+      "step": 700
+    },
+    {
+      "epoch": 0.20235285639649053,
+      "grad_norm": 1.7175828218460083,
+      "learning_rate": 8.397572078907437e-06,
+      "loss": 0.2734,
+      "step": 702
+    },
+    {
+      "epoch": 0.20321761219305673,
+      "grad_norm": 1.8725277185440063,
+      "learning_rate": 8.38846737481032e-06,
+      "loss": 0.2309,
+      "step": 705
+    },
+    {
+      "epoch": 0.20408236798962293,
+      "grad_norm": 1.7434577941894531,
+      "learning_rate": 8.379362670713203e-06,
+      "loss": 0.2395,
+      "step": 708
+    },
+    {
+      "epoch": 0.20494712378618912,
+      "grad_norm": 2.8038480281829834,
+      "learning_rate": 8.370257966616086e-06,
+      "loss": 0.2787,
+      "step": 711
+    },
+    {
+      "epoch": 0.20581187958275532,
+      "grad_norm": 1.701920509338379,
+      "learning_rate": 8.361153262518968e-06,
+      "loss": 0.2267,
+      "step": 714
+    },
+    {
+      "epoch": 0.20667663537932152,
+      "grad_norm": 4.903564453125,
+      "learning_rate": 8.352048558421853e-06,
+      "loss": 0.2716,
+      "step": 717
+    },
+    {
+      "epoch": 0.2075413911758877,
+      "grad_norm": 3.723651647567749,
+      "learning_rate": 8.342943854324736e-06,
+      "loss": 0.2674,
+      "step": 720
+    },
+    {
+      "epoch": 0.20840614697245394,
+      "grad_norm": 1.7158082723617554,
+      "learning_rate": 8.333839150227619e-06,
+      "loss": 0.2296,
+      "step": 723
+    },
+    {
+      "epoch": 0.20927090276902013,
+      "grad_norm": 2.1699960231781006,
+      "learning_rate": 8.324734446130501e-06,
+      "loss": 0.2843,
+      "step": 726
+    },
+    {
+      "epoch": 0.21013565856558633,
+      "grad_norm": 4.244576454162598,
+      "learning_rate": 8.315629742033384e-06,
+      "loss": 0.2678,
+      "step": 729
+    },
+    {
+      "epoch": 0.21100041436215253,
+      "grad_norm": 1.4597645998001099,
+      "learning_rate": 8.306525037936269e-06,
+      "loss": 0.248,
+      "step": 732
+    },
+    {
+      "epoch": 0.21186517015871872,
+      "grad_norm": 2.9361813068389893,
+      "learning_rate": 8.297420333839152e-06,
+      "loss": 0.241,
+      "step": 735
+    },
+    {
+      "epoch": 0.21272992595528492,
+      "grad_norm": 4.552188396453857,
+      "learning_rate": 8.288315629742034e-06,
+      "loss": 0.2935,
+      "step": 738
+    },
+    {
+      "epoch": 0.21359468175185112,
+      "grad_norm": 4.201780796051025,
+      "learning_rate": 8.279210925644917e-06,
+      "loss": 0.2935,
+      "step": 741
+    },
+    {
+      "epoch": 0.21445943754841731,
+      "grad_norm": 7.236446380615234,
+      "learning_rate": 8.2701062215478e-06,
+      "loss": 0.2584,
+      "step": 744
+    },
+    {
+      "epoch": 0.2153241933449835,
+      "grad_norm": 1.7462209463119507,
+      "learning_rate": 8.261001517450683e-06,
+      "loss": 0.2457,
+      "step": 747
+    },
+    {
+      "epoch": 0.2161889491415497,
+      "grad_norm": 3.9612677097320557,
+      "learning_rate": 8.251896813353568e-06,
+      "loss": 0.2472,
+      "step": 750
+    },
+    {
+      "epoch": 0.2170537049381159,
+      "grad_norm": 3.581313371658325,
+      "learning_rate": 8.24279210925645e-06,
+      "loss": 0.2833,
+      "step": 753
+    },
+    {
+      "epoch": 0.2179184607346821,
+      "grad_norm": 3.473001003265381,
+      "learning_rate": 8.233687405159333e-06,
+      "loss": 0.2474,
+      "step": 756
+    },
+    {
+      "epoch": 0.2187832165312483,
+      "grad_norm": 5.877528667449951,
+      "learning_rate": 8.224582701062216e-06,
+      "loss": 0.2469,
+      "step": 759
+    },
+    {
+      "epoch": 0.21964797232781452,
+      "grad_norm": 4.294084072113037,
+      "learning_rate": 8.215477996965099e-06,
+      "loss": 0.2569,
+      "step": 762
+    },
+    {
+      "epoch": 0.22051272812438072,
+      "grad_norm": 1.9812819957733154,
+      "learning_rate": 8.206373292867983e-06,
+      "loss": 0.2592,
+      "step": 765
+    },
+    {
+      "epoch": 0.22137748392094692,
+      "grad_norm": 1.6627389192581177,
+      "learning_rate": 8.197268588770866e-06,
+      "loss": 0.2597,
+      "step": 768
+    },
+    {
+      "epoch": 0.2222422397175131,
+      "grad_norm": 2.557081699371338,
+      "learning_rate": 8.188163884673749e-06,
+      "loss": 0.2624,
+      "step": 771
+    },
+    {
+      "epoch": 0.2231069955140793,
+      "grad_norm": 3.3301448822021484,
+      "learning_rate": 8.179059180576632e-06,
+      "loss": 0.2458,
+      "step": 774
+    },
+    {
+      "epoch": 0.2239717513106455,
+      "grad_norm": 3.717036247253418,
+      "learning_rate": 8.169954476479515e-06,
+      "loss": 0.2403,
+      "step": 777
+    },
+    {
+      "epoch": 0.2248365071072117,
+      "grad_norm": 1.9032166004180908,
+      "learning_rate": 8.1608497723824e-06,
+      "loss": 0.2194,
+      "step": 780
+    },
+    {
+      "epoch": 0.2257012629037779,
+      "grad_norm": 11.293305397033691,
+      "learning_rate": 8.151745068285282e-06,
+      "loss": 0.276,
+      "step": 783
+    },
+    {
+      "epoch": 0.2265660187003441,
+      "grad_norm": 2.2903361320495605,
+      "learning_rate": 8.142640364188165e-06,
+      "loss": 0.2089,
+      "step": 786
+    },
+    {
+      "epoch": 0.2274307744969103,
+      "grad_norm": 1.6450647115707397,
+      "learning_rate": 8.133535660091048e-06,
+      "loss": 0.239,
+      "step": 789
+    },
+    {
+      "epoch": 0.2282955302934765,
+      "grad_norm": 2.290724277496338,
+      "learning_rate": 8.12443095599393e-06,
+      "loss": 0.2509,
+      "step": 792
+    },
+    {
+      "epoch": 0.2291602860900427,
+      "grad_norm": 2.6252450942993164,
+      "learning_rate": 8.115326251896815e-06,
+      "loss": 0.2578,
+      "step": 795
+    },
+    {
+      "epoch": 0.23002504188660888,
+      "grad_norm": 4.217925071716309,
+      "learning_rate": 8.106221547799698e-06,
+      "loss": 0.2604,
+      "step": 798
+    },
+    {
+      "epoch": 0.23060154575098638,
+      "eval_loss": 0.23178604245185852,
+      "eval_mse": 0.2317860303344205,
+      "eval_runtime": 6.6317,
+      "eval_samples_per_second": 150.79,
+      "eval_steps_per_second": 18.849,
+      "step": 800
+    },
+    {
+      "epoch": 0.2308897976831751,
+      "grad_norm": 2.014478921890259,
+      "learning_rate": 8.097116843702581e-06,
+      "loss": 0.238,
+      "step": 801
+    },
+    {
+      "epoch": 0.2317545534797413,
+      "grad_norm": 3.1536102294921875,
+      "learning_rate": 8.088012139605464e-06,
+      "loss": 0.2236,
+      "step": 804
+    },
+    {
+      "epoch": 0.2326193092763075,
+      "grad_norm": 2.094320297241211,
+      "learning_rate": 8.078907435508347e-06,
+      "loss": 0.2536,
+      "step": 807
+    },
+    {
+      "epoch": 0.2334840650728737,
+      "grad_norm": 1.7041524648666382,
+      "learning_rate": 8.06980273141123e-06,
+      "loss": 0.2243,
+      "step": 810
+    },
+    {
+      "epoch": 0.2343488208694399,
+      "grad_norm": 4.586849689483643,
+      "learning_rate": 8.060698027314114e-06,
+      "loss": 0.2276,
+      "step": 813
+    },
+    {
+      "epoch": 0.2352135766660061,
+      "grad_norm": 1.8360718488693237,
+      "learning_rate": 8.051593323216997e-06,
+      "loss": 0.2299,
+      "step": 816
+    },
+    {
+      "epoch": 0.2360783324625723,
+      "grad_norm": 2.283907651901245,
+      "learning_rate": 8.04248861911988e-06,
+      "loss": 0.2996,
+      "step": 819
+    },
+    {
+      "epoch": 0.23694308825913848,
+      "grad_norm": 2.762160301208496,
+      "learning_rate": 8.033383915022763e-06,
+      "loss": 0.2474,
+      "step": 822
+    },
+    {
+      "epoch": 0.23780784405570468,
+      "grad_norm": 2.776780366897583,
+      "learning_rate": 8.024279210925645e-06,
+      "loss": 0.2394,
+      "step": 825
+    },
+    {
+      "epoch": 0.23867259985227088,
+      "grad_norm": 1.8678719997406006,
+      "learning_rate": 8.01517450682853e-06,
+      "loss": 0.2732,
+      "step": 828
+    },
+    {
+      "epoch": 0.23953735564883707,
+      "grad_norm": 1.6664822101593018,
+      "learning_rate": 8.006069802731413e-06,
+      "loss": 0.245,
+      "step": 831
+    },
+    {
+      "epoch": 0.24040211144540327,
+      "grad_norm": 2.244666814804077,
+      "learning_rate": 7.996965098634296e-06,
+      "loss": 0.2587,
+      "step": 834
+    },
+    {
+      "epoch": 0.24126686724196947,
+      "grad_norm": 1.8958755731582642,
+      "learning_rate": 7.987860394537178e-06,
+      "loss": 0.2632,
+      "step": 837
+    },
+    {
+      "epoch": 0.2421316230385357,
+      "grad_norm": 1.7408092021942139,
+      "learning_rate": 7.978755690440061e-06,
+      "loss": 0.2475,
+      "step": 840
+    },
+    {
+      "epoch": 0.2429963788351019,
+      "grad_norm": 4.280789375305176,
+      "learning_rate": 7.969650986342944e-06,
+      "loss": 0.2478,
+      "step": 843
+    },
+    {
+      "epoch": 0.2438611346316681,
+      "grad_norm": 3.1445536613464355,
+      "learning_rate": 7.960546282245829e-06,
+      "loss": 0.2412,
+      "step": 846
+    },
+    {
+      "epoch": 0.24472589042823428,
+      "grad_norm": 1.806154727935791,
+      "learning_rate": 7.951441578148712e-06,
+      "loss": 0.2772,
+      "step": 849
+    },
+    {
+      "epoch": 0.24559064622480048,
+      "grad_norm": 2.123932123184204,
+      "learning_rate": 7.942336874051594e-06,
+      "loss": 0.2535,
+      "step": 852
+    },
+    {
+      "epoch": 0.24645540202136668,
+      "grad_norm": 1.488558292388916,
+      "learning_rate": 7.933232169954477e-06,
+      "loss": 0.2342,
+      "step": 855
+    },
+    {
+      "epoch": 0.24732015781793287,
+      "grad_norm": 3.2414135932922363,
+      "learning_rate": 7.92412746585736e-06,
+      "loss": 0.2536,
+      "step": 858
+    },
+    {
+      "epoch": 0.24818491361449907,
+      "grad_norm": 1.952009916305542,
+      "learning_rate": 7.915022761760245e-06,
+      "loss": 0.2028,
+      "step": 861
+    },
+    {
+      "epoch": 0.24904966941106527,
+      "grad_norm": 1.7001383304595947,
+      "learning_rate": 7.905918057663127e-06,
+      "loss": 0.2791,
+      "step": 864
+    },
+    {
+      "epoch": 0.24991442520763146,
+      "grad_norm": 2.1715829372406006,
+      "learning_rate": 7.89681335356601e-06,
+      "loss": 0.2211,
+      "step": 867
+    },
+    {
+      "epoch": 0.25077918100419766,
+      "grad_norm": 2.4868202209472656,
+      "learning_rate": 7.887708649468893e-06,
+      "loss": 0.2599,
+      "step": 870
+    },
+    {
+      "epoch": 0.25164393680076386,
+      "grad_norm": 1.4740101099014282,
+      "learning_rate": 7.878603945371776e-06,
+      "loss": 0.2076,
+      "step": 873
+    },
+    {
+      "epoch": 0.25250869259733005,
+      "grad_norm": 2.241642951965332,
+      "learning_rate": 7.869499241274659e-06,
+      "loss": 0.2413,
+      "step": 876
+    },
+    {
+      "epoch": 0.25337344839389625,
+      "grad_norm": 4.3130784034729,
+      "learning_rate": 7.860394537177543e-06,
+      "loss": 0.2495,
+      "step": 879
+    },
+    {
+      "epoch": 0.25423820419046245,
+      "grad_norm": 6.674787998199463,
+      "learning_rate": 7.851289833080426e-06,
+      "loss": 0.2396,
+      "step": 882
+    },
+    {
+      "epoch": 0.25510295998702864,
+      "grad_norm": 2.463395118713379,
+      "learning_rate": 7.842185128983309e-06,
+      "loss": 0.2304,
+      "step": 885
+    },
+    {
+      "epoch": 0.25596771578359484,
+      "grad_norm": 4.519803047180176,
+      "learning_rate": 7.833080424886192e-06,
+      "loss": 0.2603,
+      "step": 888
+    },
+    {
+      "epoch": 0.25683247158016104,
+      "grad_norm": 1.3876014947891235,
+      "learning_rate": 7.823975720789075e-06,
+      "loss": 0.2282,
+      "step": 891
+    },
+    {
+      "epoch": 0.25769722737672723,
+      "grad_norm": 2.0575900077819824,
+      "learning_rate": 7.814871016691958e-06,
+      "loss": 0.2286,
+      "step": 894
+    },
+    {
+      "epoch": 0.2585619831732935,
+      "grad_norm": 4.538529872894287,
+      "learning_rate": 7.805766312594842e-06,
+      "loss": 0.2606,
+      "step": 897
+    },
+    {
+      "epoch": 0.2594267389698597,
+      "grad_norm": 3.036686658859253,
+      "learning_rate": 7.796661608497725e-06,
+      "loss": 0.2961,
+      "step": 900
+    },
+    {
+      "epoch": 0.2594267389698597,
+      "eval_loss": 0.2185884416103363,
+      "eval_mse": 0.21858844196051358,
+      "eval_runtime": 6.596,
+      "eval_samples_per_second": 151.607,
+      "eval_steps_per_second": 18.951,
+      "step": 900
+    },
+    {
+      "epoch": 0.2602914947664259,
+      "grad_norm": 2.319840669631958,
+      "learning_rate": 7.787556904400608e-06,
+      "loss": 0.2493,
+      "step": 903
+    },
+    {
+      "epoch": 0.2611562505629921,
+      "grad_norm": 4.723567485809326,
+      "learning_rate": 7.77845220030349e-06,
+      "loss": 0.2236,
+      "step": 906
+    },
+    {
+      "epoch": 0.2620210063595583,
+      "grad_norm": 2.231250286102295,
+      "learning_rate": 7.769347496206373e-06,
+      "loss": 0.2766,
+      "step": 909
+    },
+    {
+      "epoch": 0.26288576215612447,
+      "grad_norm": 2.3697762489318848,
+      "learning_rate": 7.760242792109256e-06,
+      "loss": 0.2319,
+      "step": 912
+    },
+    {
+      "epoch": 0.26375051795269067,
+      "grad_norm": 2.0771355628967285,
+      "learning_rate": 7.75113808801214e-06,
+      "loss": 0.246,
+      "step": 915
+    },
+    {
+      "epoch": 0.26461527374925686,
+      "grad_norm": 2.4788358211517334,
+      "learning_rate": 7.742033383915024e-06,
+      "loss": 0.2163,
+      "step": 918
+    },
+    {
+      "epoch": 0.26548002954582306,
+      "grad_norm": 1.5741719007492065,
+      "learning_rate": 7.732928679817907e-06,
+      "loss": 0.2308,
+      "step": 921
+    },
+    {
+      "epoch": 0.26634478534238926,
+      "grad_norm": 2.12677001953125,
+      "learning_rate": 7.72382397572079e-06,
+      "loss": 0.2304,
+      "step": 924
+    },
+    {
+      "epoch": 0.26720954113895545,
+      "grad_norm": 1.9022995233535767,
+      "learning_rate": 7.714719271623672e-06,
+      "loss": 0.2454,
+      "step": 927
+    },
+    {
+      "epoch": 0.26807429693552165,
+      "grad_norm": 3.3529253005981445,
+      "learning_rate": 7.705614567526557e-06,
+      "loss": 0.2137,
+      "step": 930
+    },
+    {
+      "epoch": 0.26893905273208785,
+      "grad_norm": 1.935281753540039,
+      "learning_rate": 7.69650986342944e-06,
+      "loss": 0.2581,
+      "step": 933
+    },
+    {
+      "epoch": 0.26980380852865404,
+      "grad_norm": 2.14315128326416,
+      "learning_rate": 7.687405159332322e-06,
+      "loss": 0.2317,
+      "step": 936
+    },
+    {
+      "epoch": 0.27066856432522024,
+      "grad_norm": 2.028090238571167,
+      "learning_rate": 7.678300455235205e-06,
+      "loss": 0.2379,
+      "step": 939
+    },
+    {
+      "epoch": 0.27153332012178644,
+      "grad_norm": 2.918959379196167,
+      "learning_rate": 7.669195751138088e-06,
+      "loss": 0.2708,
+      "step": 942
+    },
+    {
+      "epoch": 0.27239807591835263,
+      "grad_norm": 4.042644500732422,
+      "learning_rate": 7.660091047040971e-06,
+      "loss": 0.3112,
+      "step": 945
+    },
+    {
+      "epoch": 0.27326283171491883,
+      "grad_norm": 1.8345041275024414,
+      "learning_rate": 7.650986342943855e-06,
+      "loss": 0.267,
+      "step": 948
+    },
+    {
+      "epoch": 0.274127587511485,
+      "grad_norm": 5.901050567626953,
+      "learning_rate": 7.641881638846738e-06,
+      "loss": 0.272,
+      "step": 951
+    },
+    {
+      "epoch": 0.2749923433080512,
+      "grad_norm": 2.413675546646118,
+      "learning_rate": 7.632776934749621e-06,
+      "loss": 0.2906,
+      "step": 954
+    },
+    {
+      "epoch": 0.2758570991046174,
+      "grad_norm": 2.699126958847046,
+      "learning_rate": 7.623672230652505e-06,
+      "loss": 0.2201,
+      "step": 957
+    },
+    {
+      "epoch": 0.2767218549011836,
+      "grad_norm": 1.3186564445495605,
+      "learning_rate": 7.614567526555388e-06,
+      "loss": 0.227,
+      "step": 960
+    },
+    {
+      "epoch": 0.2775866106977498,
+      "grad_norm": 7.576478958129883,
+      "learning_rate": 7.6054628224582705e-06,
+      "loss": 0.2569,
+      "step": 963
+    },
+    {
+      "epoch": 0.278451366494316,
+      "grad_norm": 1.853145718574524,
+      "learning_rate": 7.596358118361153e-06,
+      "loss": 0.2586,
+      "step": 966
+    },
+    {
+      "epoch": 0.2793161222908822,
+      "grad_norm": 3.2667322158813477,
+      "learning_rate": 7.587253414264037e-06,
+      "loss": 0.2483,
+      "step": 969
+    },
+    {
+      "epoch": 0.2801808780874484,
+      "grad_norm": 2.7756762504577637,
+      "learning_rate": 7.578148710166921e-06,
+      "loss": 0.2724,
+      "step": 972
+    },
+    {
+      "epoch": 0.28104563388401466,
+      "grad_norm": 2.138936758041382,
+      "learning_rate": 7.569044006069804e-06,
+      "loss": 0.2517,
+      "step": 975
+    },
+    {
+      "epoch": 0.28191038968058085,
+      "grad_norm": 1.9090849161148071,
+      "learning_rate": 7.5599393019726864e-06,
+      "loss": 0.2608,
+      "step": 978
+    },
+    {
+      "epoch": 0.28277514547714705,
+      "grad_norm": 2.6184728145599365,
+      "learning_rate": 7.550834597875569e-06,
+      "loss": 0.2728,
+      "step": 981
+    },
+    {
+      "epoch": 0.28363990127371325,
+      "grad_norm": 1.5910395383834839,
+      "learning_rate": 7.541729893778453e-06,
+      "loss": 0.2144,
+      "step": 984
+    },
+    {
+      "epoch": 0.28450465707027944,
+      "grad_norm": 2.3524558544158936,
+      "learning_rate": 7.532625189681337e-06,
+      "loss": 0.23,
+      "step": 987
+    },
+    {
+      "epoch": 0.28536941286684564,
+      "grad_norm": 3.056361675262451,
+      "learning_rate": 7.5235204855842195e-06,
+      "loss": 0.2397,
+      "step": 990
+    },
+    {
+      "epoch": 0.28623416866341184,
+      "grad_norm": 3.4158847332000732,
+      "learning_rate": 7.514415781487102e-06,
+      "loss": 0.2712,
+      "step": 993
+    },
+    {
+      "epoch": 0.28709892445997803,
+      "grad_norm": 3.3620333671569824,
+      "learning_rate": 7.505311077389985e-06,
+      "loss": 0.2107,
+      "step": 996
+    },
+    {
+      "epoch": 0.28796368025654423,
+      "grad_norm": 3.508890390396118,
+      "learning_rate": 7.496206373292868e-06,
+      "loss": 0.2453,
+      "step": 999
+    },
+    {
+      "epoch": 0.28825193218873296,
+      "eval_loss": 0.21684841811656952,
+      "eval_mse": 0.21684841979760677,
+      "eval_runtime": 6.6234,
+      "eval_samples_per_second": 150.98,
+      "eval_steps_per_second": 18.872,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2888284360531104,
+      "grad_norm": 2.658740997314453,
+      "learning_rate": 7.487101669195752e-06,
+      "loss": 0.2319,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2896931918496766,
+      "grad_norm": 1.370540976524353,
+      "learning_rate": 7.477996965098635e-06,
+      "loss": 0.2443,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2905579476462428,
+      "grad_norm": 8.996585845947266,
+      "learning_rate": 7.468892261001518e-06,
+      "loss": 0.2859,
+      "step": 1008
+    },
+    {
+      "epoch": 0.291422703442809,
+      "grad_norm": 3.880053758621216,
+      "learning_rate": 7.459787556904401e-06,
+      "loss": 0.2913,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2922874592393752,
+      "grad_norm": 4.735537052154541,
+      "learning_rate": 7.450682852807284e-06,
+      "loss": 0.258,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2931522150359414,
+      "grad_norm": 1.7119396924972534,
+      "learning_rate": 7.441578148710168e-06,
+      "loss": 0.2397,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2940169708325076,
+      "grad_norm": 3.4769861698150635,
+      "learning_rate": 7.4324734446130505e-06,
+      "loss": 0.209,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2948817266290738,
+      "grad_norm": 2.3741278648376465,
+      "learning_rate": 7.423368740515934e-06,
+      "loss": 0.2627,
+      "step": 1023
+    },
+    {
+      "epoch": 0.29574648242564,
+      "grad_norm": 1.5303798913955688,
+      "learning_rate": 7.414264036418817e-06,
+      "loss": 0.2352,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2966112382222062,
+      "grad_norm": 1.63661789894104,
+      "learning_rate": 7.4051593323217e-06,
+      "loss": 0.2448,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2974759940187724,
+      "grad_norm": 1.531538724899292,
+      "learning_rate": 7.3960546282245835e-06,
+      "loss": 0.23,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2983407498153386,
+      "grad_norm": 1.3936281204223633,
+      "learning_rate": 7.386949924127466e-06,
+      "loss": 0.2367,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2992055056119048,
+      "grad_norm": 4.795119762420654,
+      "learning_rate": 7.377845220030349e-06,
+      "loss": 0.2621,
+      "step": 1038
+    },
+    {
+      "epoch": 0.300070261408471,
+      "grad_norm": 3.725170135498047,
+      "learning_rate": 7.368740515933233e-06,
+      "loss": 0.2708,
+      "step": 1041
+    },
+    {
+      "epoch": 0.3009350172050372,
+      "grad_norm": 1.5772522687911987,
+      "learning_rate": 7.359635811836116e-06,
+      "loss": 0.1897,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3017997730016034,
+      "grad_norm": 4.716231822967529,
+      "learning_rate": 7.3505311077389994e-06,
+      "loss": 0.2297,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3026645287981696,
+      "grad_norm": 3.9254777431488037,
+      "learning_rate": 7.341426403641882e-06,
+      "loss": 0.2285,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3035292845947358,
+      "grad_norm": 2.7238216400146484,
+      "learning_rate": 7.332321699544765e-06,
+      "loss": 0.2488,
+      "step": 1053
+    },
+    {
+      "epoch": 0.304394040391302,
+      "grad_norm": 7.061728477478027,
+      "learning_rate": 7.323216995447649e-06,
+      "loss": 0.2329,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3052587961878682,
+      "grad_norm": 2.2323367595672607,
+      "learning_rate": 7.314112291350532e-06,
+      "loss": 0.2511,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3061235519844344,
+      "grad_norm": 3.432279586791992,
+      "learning_rate": 7.305007587253415e-06,
+      "loss": 0.251,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3069883077810006,
+      "grad_norm": 4.372462272644043,
+      "learning_rate": 7.295902883156298e-06,
+      "loss": 0.2306,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3078530635775668,
+      "grad_norm": 3.428677558898926,
+      "learning_rate": 7.286798179059181e-06,
+      "loss": 0.2389,
+      "step": 1068
+    },
+    {
+      "epoch": 0.308717819374133,
+      "grad_norm": 1.7803095579147339,
+      "learning_rate": 7.277693474962064e-06,
+      "loss": 0.2374,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3095825751706992,
+      "grad_norm": 1.7453327178955078,
+      "learning_rate": 7.2685887708649476e-06,
+      "loss": 0.2175,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3104473309672654,
+      "grad_norm": 3.4764058589935303,
+      "learning_rate": 7.25948406676783e-06,
+      "loss": 0.2649,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3113120867638316,
+      "grad_norm": 2.5065643787384033,
+      "learning_rate": 7.250379362670714e-06,
+      "loss": 0.2436,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3121768425603978,
+      "grad_norm": 2.4735498428344727,
+      "learning_rate": 7.241274658573597e-06,
+      "loss": 0.2565,
+      "step": 1083
+    },
+    {
+      "epoch": 0.313041598356964,
+      "grad_norm": 7.107683181762695,
+      "learning_rate": 7.23216995447648e-06,
+      "loss": 0.2691,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3139063541535302,
+      "grad_norm": 1.9517550468444824,
+      "learning_rate": 7.223065250379363e-06,
+      "loss": 0.2869,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3147711099500964,
+      "grad_norm": 3.1179702281951904,
+      "learning_rate": 7.213960546282246e-06,
+      "loss": 0.2434,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3156358657466626,
+      "grad_norm": 4.619999885559082,
+      "learning_rate": 7.20485584218513e-06,
+      "loss": 0.2469,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3165006215432288,
+      "grad_norm": 6.7724528312683105,
+      "learning_rate": 7.195751138088013e-06,
+      "loss": 0.278,
+      "step": 1098
+    },
+    {
+      "epoch": 0.31707712540760624,
+      "eval_loss": 0.2247343808412552,
+      "eval_mse": 0.22473438137583435,
+      "eval_runtime": 6.5256,
+      "eval_samples_per_second": 153.242,
+      "eval_steps_per_second": 19.155,
+      "step": 1100
+    },
+    {
+      "epoch": 0.317365377339795,
+      "grad_norm": 1.6179205179214478,
+      "learning_rate": 7.186646433990896e-06,
+      "loss": 0.252,
+      "step": 1101
+    },
+    {
+      "epoch": 0.31823013313636117,
+      "grad_norm": 1.8803197145462036,
+      "learning_rate": 7.1775417298937785e-06,
+      "loss": 0.2227,
+      "step": 1104
+    },
+    {
+      "epoch": 0.31909488893292737,
+      "grad_norm": 2.2427573204040527,
+      "learning_rate": 7.168437025796661e-06,
+      "loss": 0.2399,
+      "step": 1107
+    },
+    {
+      "epoch": 0.31995964472949356,
+      "grad_norm": 1.907244086265564,
+      "learning_rate": 7.159332321699546e-06,
+      "loss": 0.2162,
+      "step": 1110
+    },
+    {
+      "epoch": 0.32082440052605976,
+      "grad_norm": 3.7878000736236572,
+      "learning_rate": 7.150227617602429e-06,
+      "loss": 0.246,
+      "step": 1113
+    },
+    {
+      "epoch": 0.32168915632262596,
+      "grad_norm": 1.9053196907043457,
+      "learning_rate": 7.141122913505312e-06,
+      "loss": 0.2684,
+      "step": 1116
+    },
+    {
+      "epoch": 0.32255391211919215,
+      "grad_norm": 5.108983039855957,
+      "learning_rate": 7.1320182094081944e-06,
+      "loss": 0.2237,
+      "step": 1119
+    },
+    {
+      "epoch": 0.32341866791575835,
+      "grad_norm": 2.2469422817230225,
+      "learning_rate": 7.122913505311077e-06,
+      "loss": 0.2462,
+      "step": 1122
+    },
+    {
+      "epoch": 0.32428342371232455,
+      "grad_norm": 5.127351760864258,
+      "learning_rate": 7.113808801213962e-06,
+      "loss": 0.2543,
+      "step": 1125
+    },
+    {
+      "epoch": 0.32514817950889074,
+      "grad_norm": 4.980170249938965,
+      "learning_rate": 7.104704097116845e-06,
+      "loss": 0.2775,
+      "step": 1128
+    },
+    {
+      "epoch": 0.326012935305457,
+      "grad_norm": 3.701903820037842,
+      "learning_rate": 7.0955993930197275e-06,
+      "loss": 0.2548,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3268776911020232,
+      "grad_norm": 3.780144214630127,
+      "learning_rate": 7.08649468892261e-06,
+      "loss": 0.2687,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3277424468985894,
+      "grad_norm": 2.2161099910736084,
+      "learning_rate": 7.077389984825493e-06,
+      "loss": 0.2351,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3286072026951556,
+      "grad_norm": 4.7017998695373535,
+      "learning_rate": 7.068285280728376e-06,
+      "loss": 0.2431,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3294719584917218,
+      "grad_norm": 2.053750991821289,
+      "learning_rate": 7.0591805766312606e-06,
+      "loss": 0.254,
+      "step": 1143
+    },
+    {
+      "epoch": 0.330336714288288,
+      "grad_norm": 2.8078341484069824,
+      "learning_rate": 7.050075872534143e-06,
+      "loss": 0.2646,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3312014700848542,
+      "grad_norm": 2.585087776184082,
+      "learning_rate": 7.040971168437026e-06,
+      "loss": 0.2536,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3320662258814204,
+      "grad_norm": 4.2963104248046875,
+      "learning_rate": 7.031866464339909e-06,
+      "loss": 0.2276,
+      "step": 1152
+    },
+    {
+      "epoch": 0.33293098167798657,
+      "grad_norm": 4.205751419067383,
+      "learning_rate": 7.022761760242792e-06,
+      "loss": 0.2404,
+      "step": 1155
+    },
+    {
+      "epoch": 0.33379573747455277,
+      "grad_norm": 3.460796356201172,
+      "learning_rate": 7.0136570561456765e-06,
+      "loss": 0.2473,
+      "step": 1158
+    },
+    {
+      "epoch": 0.33466049327111896,
+      "grad_norm": 3.529181957244873,
+      "learning_rate": 7.004552352048559e-06,
+      "loss": 0.24,
+      "step": 1161
+    },
+    {
+      "epoch": 0.33552524906768516,
+      "grad_norm": 3.203437328338623,
+      "learning_rate": 6.995447647951442e-06,
+      "loss": 0.2547,
+      "step": 1164
+    },
+    {
+      "epoch": 0.33639000486425136,
+      "grad_norm": 4.1535491943359375,
+      "learning_rate": 6.986342943854325e-06,
+      "loss": 0.2532,
+      "step": 1167
+    },
+    {
+      "epoch": 0.33725476066081755,
+      "grad_norm": 3.7933478355407715,
+      "learning_rate": 6.977238239757208e-06,
+      "loss": 0.2435,
+      "step": 1170
+    },
+    {
+      "epoch": 0.33811951645738375,
+      "grad_norm": 2.6703147888183594,
+      "learning_rate": 6.968133535660092e-06,
+      "loss": 0.2293,
+      "step": 1173
+    },
+    {
+      "epoch": 0.33898427225394995,
+      "grad_norm": 2.8900182247161865,
+      "learning_rate": 6.959028831562975e-06,
+      "loss": 0.2163,
+      "step": 1176
+    },
+    {
+      "epoch": 0.33984902805051614,
+      "grad_norm": 3.9563350677490234,
+      "learning_rate": 6.949924127465858e-06,
+      "loss": 0.2682,
+      "step": 1179
+    },
+    {
+      "epoch": 0.34071378384708234,
+      "grad_norm": 1.8461941480636597,
+      "learning_rate": 6.940819423368741e-06,
+      "loss": 0.2293,
+      "step": 1182
+    },
+    {
+      "epoch": 0.34157853964364854,
+      "grad_norm": 3.313368082046509,
+      "learning_rate": 6.931714719271624e-06,
+      "loss": 0.2436,
+      "step": 1185
+    },
+    {
+      "epoch": 0.34244329544021473,
+      "grad_norm": 1.7820873260498047,
+      "learning_rate": 6.922610015174508e-06,
+      "loss": 0.212,
+      "step": 1188
+    },
+    {
+      "epoch": 0.34330805123678093,
+      "grad_norm": 1.995291829109192,
+      "learning_rate": 6.913505311077391e-06,
+      "loss": 0.2243,
+      "step": 1191
+    },
+    {
+      "epoch": 0.34417280703334713,
+      "grad_norm": 2.928727626800537,
+      "learning_rate": 6.904400606980274e-06,
+      "loss": 0.2296,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3450375628299133,
+      "grad_norm": 2.5598068237304688,
+      "learning_rate": 6.895295902883157e-06,
+      "loss": 0.2135,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3459023186264795,
+      "grad_norm": 2.4700326919555664,
+      "learning_rate": 6.88619119878604e-06,
+      "loss": 0.2319,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3459023186264795,
+      "eval_loss": 0.21415139734745026,
+      "eval_mse": 0.2141513990436215,
+      "eval_runtime": 6.5018,
+      "eval_samples_per_second": 153.803,
+      "eval_steps_per_second": 19.225,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3467670744230457,
+      "grad_norm": 3.6825876235961914,
+      "learning_rate": 6.8770864946889225e-06,
+      "loss": 0.2409,
+      "step": 1203
+    },
+    {
+      "epoch": 0.3476318302196119,
+      "grad_norm": 2.8780245780944824,
+      "learning_rate": 6.867981790591807e-06,
+      "loss": 0.2219,
+      "step": 1206
+    },
+    {
+      "epoch": 0.34849658601617817,
+      "grad_norm": 2.641505479812622,
+      "learning_rate": 6.85887708649469e-06,
+      "loss": 0.2315,
+      "step": 1209
+    },
+    {
+      "epoch": 0.34936134181274436,
+      "grad_norm": 1.660909652709961,
+      "learning_rate": 6.849772382397573e-06,
+      "loss": 0.224,
+      "step": 1212
+    },
+    {
+      "epoch": 0.35022609760931056,
+      "grad_norm": 5.104085445404053,
+      "learning_rate": 6.8406676783004556e-06,
+      "loss": 0.2184,
+      "step": 1215
+    },
+    {
+      "epoch": 0.35109085340587676,
+      "grad_norm": 1.870938777923584,
+      "learning_rate": 6.831562974203338e-06,
+      "loss": 0.234,
+      "step": 1218
+    },
+    {
+      "epoch": 0.35195560920244295,
+      "grad_norm": 4.542322635650635,
+      "learning_rate": 6.822458270106223e-06,
+      "loss": 0.2247,
+      "step": 1221
+    },
+    {
+      "epoch": 0.35282036499900915,
+      "grad_norm": 1.4236701726913452,
+      "learning_rate": 6.813353566009106e-06,
+      "loss": 0.2168,
+      "step": 1224
+    },
+    {
+      "epoch": 0.35368512079557535,
+      "grad_norm": 2.0418410301208496,
+      "learning_rate": 6.804248861911989e-06,
+      "loss": 0.2425,
+      "step": 1227
+    },
+    {
+      "epoch": 0.35454987659214154,
+      "grad_norm": 2.868399143218994,
+      "learning_rate": 6.7951441578148715e-06,
+      "loss": 0.2354,
+      "step": 1230
+    },
+    {
+      "epoch": 0.35541463238870774,
+      "grad_norm": 2.453361749649048,
+      "learning_rate": 6.786039453717754e-06,
+      "loss": 0.2266,
+      "step": 1233
+    },
+    {
+      "epoch": 0.35627938818527394,
+      "grad_norm": 2.0826542377471924,
+      "learning_rate": 6.776934749620638e-06,
+      "loss": 0.2194,
+      "step": 1236
+    },
+    {
+      "epoch": 0.35714414398184013,
+      "grad_norm": 1.511440634727478,
+      "learning_rate": 6.767830045523522e-06,
+      "loss": 0.2264,
+      "step": 1239
+    },
+    {
+      "epoch": 0.35800889977840633,
+      "grad_norm": 2.456897020339966,
+      "learning_rate": 6.7587253414264045e-06,
+      "loss": 0.2864,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3588736555749725,
+      "grad_norm": 2.873429298400879,
+      "learning_rate": 6.749620637329287e-06,
+      "loss": 0.2055,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3597384113715387,
+      "grad_norm": 1.7501113414764404,
+      "learning_rate": 6.74051593323217e-06,
+      "loss": 0.2205,
+      "step": 1248
+    },
+    {
+      "epoch": 0.3606031671681049,
+      "grad_norm": 1.615105390548706,
+      "learning_rate": 6.731411229135054e-06,
+      "loss": 0.2337,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3614679229646711,
+      "grad_norm": 1.6872748136520386,
+      "learning_rate": 6.722306525037937e-06,
+      "loss": 0.2164,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3623326787612373,
+      "grad_norm": 3.1001312732696533,
+      "learning_rate": 6.7132018209408204e-06,
+      "loss": 0.2154,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3631974345578035,
+      "grad_norm": 2.1165292263031006,
+      "learning_rate": 6.704097116843703e-06,
+      "loss": 0.2512,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3640621903543697,
+      "grad_norm": 4.326318740844727,
+      "learning_rate": 6.694992412746586e-06,
+      "loss": 0.2471,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3649269461509359,
+      "grad_norm": 5.773290634155273,
+      "learning_rate": 6.685887708649469e-06,
+      "loss": 0.2407,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3657917019475021,
+      "grad_norm": 3.2723119258880615,
+      "learning_rate": 6.676783004552353e-06,
+      "loss": 0.3129,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3666564577440683,
+      "grad_norm": 2.927086114883423,
+      "learning_rate": 6.6676783004552355e-06,
+      "loss": 0.231,
+      "step": 1272
+    },
+    {
+      "epoch": 0.3675212135406345,
+      "grad_norm": 1.7322252988815308,
+      "learning_rate": 6.658573596358119e-06,
+      "loss": 0.2504,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3683859693372007,
+      "grad_norm": 2.5904715061187744,
+      "learning_rate": 6.649468892261002e-06,
+      "loss": 0.2079,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3692507251337669,
+      "grad_norm": 2.6561062335968018,
+      "learning_rate": 6.640364188163885e-06,
+      "loss": 0.2423,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3701154809303331,
+      "grad_norm": 3.3299241065979004,
+      "learning_rate": 6.6312594840667686e-06,
+      "loss": 0.2386,
+      "step": 1284
+    },
+    {
+      "epoch": 0.37098023672689934,
+      "grad_norm": 1.731477975845337,
+      "learning_rate": 6.622154779969651e-06,
+      "loss": 0.2017,
+      "step": 1287
+    },
+    {
+      "epoch": 0.37184499252346553,
+      "grad_norm": 2.5077965259552,
+      "learning_rate": 6.613050075872534e-06,
+      "loss": 0.2446,
+      "step": 1290
+    },
+    {
+      "epoch": 0.37270974832003173,
+      "grad_norm": 2.36556077003479,
+      "learning_rate": 6.603945371775418e-06,
+      "loss": 0.2544,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3735745041165979,
+      "grad_norm": 1.9789232015609741,
+      "learning_rate": 6.594840667678301e-06,
+      "loss": 0.2247,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3744392599131641,
+      "grad_norm": 1.7310489416122437,
+      "learning_rate": 6.5857359635811845e-06,
+      "loss": 0.1983,
+      "step": 1299
+    },
+    {
+      "epoch": 0.37472751184535286,
+      "eval_loss": 0.2175411880016327,
+      "eval_mse": 0.21754117820138344,
+      "eval_runtime": 7.1272,
+      "eval_samples_per_second": 140.308,
+      "eval_steps_per_second": 17.538,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3753040157097303,
+      "grad_norm": 4.152148723602295,
+      "learning_rate": 6.576631259484067e-06,
+      "loss": 0.2899,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3761687715062965,
+      "grad_norm": 2.1581296920776367,
+      "learning_rate": 6.56752655538695e-06,
+      "loss": 0.2448,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3770335273028627,
+      "grad_norm": 2.8600330352783203,
+      "learning_rate": 6.558421851289834e-06,
+      "loss": 0.2084,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3778982830994289,
+      "grad_norm": 1.5723686218261719,
+      "learning_rate": 6.549317147192717e-06,
+      "loss": 0.2441,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3787630388959951,
+      "grad_norm": 3.4149153232574463,
+      "learning_rate": 6.5402124430956e-06,
+      "loss": 0.247,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3796277946925613,
+      "grad_norm": 16.35414695739746,
+      "learning_rate": 6.531107738998483e-06,
+      "loss": 0.2478,
+      "step": 1317
+    },
+    {
+      "epoch": 0.3804925504891275,
+      "grad_norm": 6.227761268615723,
+      "learning_rate": 6.522003034901366e-06,
+      "loss": 0.2316,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3813573062856937,
+      "grad_norm": 2.2669637203216553,
+      "learning_rate": 6.512898330804249e-06,
+      "loss": 0.2836,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3822220620822599,
+      "grad_norm": 1.4385027885437012,
+      "learning_rate": 6.503793626707133e-06,
+      "loss": 0.2122,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3830868178788261,
+      "grad_norm": 2.3909130096435547,
+      "learning_rate": 6.4946889226100154e-06,
+      "loss": 0.2357,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3839515736753923,
+      "grad_norm": 1.7610464096069336,
+      "learning_rate": 6.485584218512899e-06,
+      "loss": 0.2746,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3848163294719585,
+      "grad_norm": 2.8983278274536133,
+      "learning_rate": 6.476479514415782e-06,
+      "loss": 0.2176,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3856810852685247,
+      "grad_norm": 1.7231597900390625,
+      "learning_rate": 6.467374810318665e-06,
+      "loss": 0.2065,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3865458410650909,
+      "grad_norm": 1.6913188695907593,
+      "learning_rate": 6.458270106221548e-06,
+      "loss": 0.2541,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3874105968616571,
+      "grad_norm": 2.1574645042419434,
+      "learning_rate": 6.449165402124431e-06,
+      "loss": 0.2393,
+      "step": 1344
+    },
+    {
+      "epoch": 0.38827535265822327,
+      "grad_norm": 3.8315231800079346,
+      "learning_rate": 6.440060698027315e-06,
+      "loss": 0.2353,
+      "step": 1347
+    },
+    {
+      "epoch": 0.38914010845478947,
+      "grad_norm": 2.655545711517334,
+      "learning_rate": 6.430955993930198e-06,
+      "loss": 0.2222,
+      "step": 1350
+    },
+    {
+      "epoch": 0.39000486425135567,
+      "grad_norm": 2.0346477031707764,
+      "learning_rate": 6.421851289833081e-06,
+      "loss": 0.2289,
+      "step": 1353
+    },
+    {
+      "epoch": 0.39086962004792186,
+      "grad_norm": 4.726449489593506,
+      "learning_rate": 6.4127465857359636e-06,
+      "loss": 0.2748,
+      "step": 1356
+    },
+    {
+      "epoch": 0.39173437584448806,
+      "grad_norm": 3.0121731758117676,
+      "learning_rate": 6.403641881638846e-06,
+      "loss": 0.2244,
+      "step": 1359
+    },
+    {
+      "epoch": 0.39259913164105426,
+      "grad_norm": 6.562178611755371,
+      "learning_rate": 6.394537177541731e-06,
+      "loss": 0.264,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3934638874376205,
+      "grad_norm": 3.1092755794525146,
+      "learning_rate": 6.385432473444614e-06,
+      "loss": 0.228,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3943286432341867,
+      "grad_norm": 4.431830406188965,
+      "learning_rate": 6.376327769347497e-06,
+      "loss": 0.2544,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3951933990307529,
+      "grad_norm": 2.545694351196289,
+      "learning_rate": 6.3672230652503795e-06,
+      "loss": 0.2618,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3960581548273191,
+      "grad_norm": 1.5483859777450562,
+      "learning_rate": 6.358118361153262e-06,
+      "loss": 0.282,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3969229106238853,
+      "grad_norm": 1.750784993171692,
+      "learning_rate": 6.349013657056147e-06,
+      "loss": 0.2447,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3977876664204515,
+      "grad_norm": 2.440020799636841,
+      "learning_rate": 6.33990895295903e-06,
+      "loss": 0.2472,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3986524222170177,
+      "grad_norm": 1.4584132432937622,
+      "learning_rate": 6.3308042488619125e-06,
+      "loss": 0.2469,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3995171780135839,
+      "grad_norm": 1.3845562934875488,
+      "learning_rate": 6.321699544764795e-06,
+      "loss": 0.2111,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4003819338101501,
+      "grad_norm": 1.736345648765564,
+      "learning_rate": 6.312594840667678e-06,
+      "loss": 0.2203,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4012466896067163,
+      "grad_norm": 2.5879809856414795,
+      "learning_rate": 6.303490136570563e-06,
+      "loss": 0.2776,
+      "step": 1392
+    },
+    {
+      "epoch": 0.4021114454032825,
+      "grad_norm": 1.3224149942398071,
+      "learning_rate": 6.294385432473446e-06,
+      "loss": 0.2016,
+      "step": 1395
+    },
+    {
+      "epoch": 0.40297620119984867,
+      "grad_norm": 2.7825825214385986,
+      "learning_rate": 6.2852807283763284e-06,
+      "loss": 0.2264,
+      "step": 1398
+    },
+    {
+      "epoch": 0.40355270506422614,
+      "eval_loss": 0.23064111173152924,
+      "eval_mse": 0.2306411211611703,
+      "eval_runtime": 6.653,
+      "eval_samples_per_second": 150.308,
+      "eval_steps_per_second": 18.789,
+      "step": 1400
+    },
+    {
+      "epoch": 0.40384095699641487,
+      "grad_norm": 3.059859275817871,
+      "learning_rate": 6.276176024279211e-06,
+      "loss": 0.1952,
+      "step": 1401
+    },
+    {
+      "epoch": 0.40470571279298106,
+      "grad_norm": 2.1431009769439697,
+      "learning_rate": 6.267071320182094e-06,
+      "loss": 0.2681,
+      "step": 1404
+    },
+    {
+      "epoch": 0.40557046858954726,
+      "grad_norm": 1.6716617345809937,
+      "learning_rate": 6.257966616084977e-06,
+      "loss": 0.221,
+      "step": 1407
+    },
+    {
+      "epoch": 0.40643522438611346,
+      "grad_norm": 1.9646525382995605,
+      "learning_rate": 6.2488619119878615e-06,
+      "loss": 0.2218,
+      "step": 1410
+    },
+    {
+      "epoch": 0.40729998018267966,
+      "grad_norm": 1.2912189960479736,
+      "learning_rate": 6.239757207890744e-06,
+      "loss": 0.2077,
+      "step": 1413
+    },
+    {
+      "epoch": 0.40816473597924585,
+      "grad_norm": 2.4723434448242188,
+      "learning_rate": 6.230652503793627e-06,
+      "loss": 0.2629,
+      "step": 1416
+    },
+    {
+      "epoch": 0.40902949177581205,
+      "grad_norm": 2.1053199768066406,
+      "learning_rate": 6.22154779969651e-06,
+      "loss": 0.2524,
+      "step": 1419
+    },
+    {
+      "epoch": 0.40989424757237825,
+      "grad_norm": 2.039580821990967,
+      "learning_rate": 6.212443095599393e-06,
+      "loss": 0.2035,
+      "step": 1422
+    },
+    {
+      "epoch": 0.41075900336894444,
+      "grad_norm": 1.499022364616394,
+      "learning_rate": 6.203338391502277e-06,
+      "loss": 0.2263,
+      "step": 1425
+    },
+    {
+      "epoch": 0.41162375916551064,
+      "grad_norm": 2.090580701828003,
+      "learning_rate": 6.19423368740516e-06,
+      "loss": 0.2834,
+      "step": 1428
+    },
+    {
+      "epoch": 0.41248851496207684,
+      "grad_norm": 2.0547232627868652,
+      "learning_rate": 6.185128983308043e-06,
+      "loss": 0.2725,
+      "step": 1431
+    },
+    {
+      "epoch": 0.41335327075864303,
+      "grad_norm": 1.8254441022872925,
+      "learning_rate": 6.176024279210926e-06,
+      "loss": 0.2234,
+      "step": 1434
+    },
+    {
+      "epoch": 0.41421802655520923,
+      "grad_norm": 1.860533595085144,
+      "learning_rate": 6.166919575113809e-06,
+      "loss": 0.2554,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4150827823517754,
+      "grad_norm": 3.225929021835327,
+      "learning_rate": 6.157814871016693e-06,
+      "loss": 0.2784,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4159475381483417,
+      "grad_norm": 2.2769436836242676,
+      "learning_rate": 6.148710166919576e-06,
+      "loss": 0.2261,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4168122939449079,
+      "grad_norm": 1.6467565298080444,
+      "learning_rate": 6.139605462822459e-06,
+      "loss": 0.2702,
+      "step": 1446
+    },
+    {
+      "epoch": 0.41767704974147407,
+      "grad_norm": 3.0362329483032227,
+      "learning_rate": 6.130500758725342e-06,
+      "loss": 0.2348,
+      "step": 1449
+    },
+    {
+      "epoch": 0.41854180553804027,
+      "grad_norm": 1.8852200508117676,
+      "learning_rate": 6.121396054628225e-06,
+      "loss": 0.2222,
+      "step": 1452
+    },
+    {
+      "epoch": 0.41940656133460646,
+      "grad_norm": 2.119568109512329,
+      "learning_rate": 6.112291350531108e-06,
+      "loss": 0.226,
+      "step": 1455
+    },
+    {
+      "epoch": 0.42027131713117266,
+      "grad_norm": 2.534950017929077,
+      "learning_rate": 6.103186646433992e-06,
+      "loss": 0.2353,
+      "step": 1458
+    },
+    {
+      "epoch": 0.42113607292773886,
+      "grad_norm": 3.6363894939422607,
+      "learning_rate": 6.094081942336875e-06,
+      "loss": 0.2724,
+      "step": 1461
+    },
+    {
+      "epoch": 0.42200082872430505,
+      "grad_norm": 1.8480486869812012,
+      "learning_rate": 6.084977238239758e-06,
+      "loss": 0.2417,
+      "step": 1464
+    },
+    {
+      "epoch": 0.42286558452087125,
+      "grad_norm": 4.110941410064697,
+      "learning_rate": 6.075872534142641e-06,
+      "loss": 0.2061,
+      "step": 1467
+    },
+    {
+      "epoch": 0.42373034031743745,
+      "grad_norm": 2.7998435497283936,
+      "learning_rate": 6.0667678300455234e-06,
+      "loss": 0.231,
+      "step": 1470
+    },
+    {
+      "epoch": 0.42459509611400365,
+      "grad_norm": 1.7628705501556396,
+      "learning_rate": 6.057663125948408e-06,
+      "loss": 0.2193,
+      "step": 1473
+    },
+    {
+      "epoch": 0.42545985191056984,
+      "grad_norm": 2.7976937294006348,
+      "learning_rate": 6.048558421851291e-06,
+      "loss": 0.2361,
+      "step": 1476
+    },
+    {
+      "epoch": 0.42632460770713604,
+      "grad_norm": 2.4593019485473633,
+      "learning_rate": 6.039453717754174e-06,
+      "loss": 0.2494,
+      "step": 1479
+    },
+    {
+      "epoch": 0.42718936350370224,
+      "grad_norm": 2.5946741104125977,
+      "learning_rate": 6.0303490136570565e-06,
+      "loss": 0.2279,
+      "step": 1482
+    },
+    {
+      "epoch": 0.42805411930026843,
+      "grad_norm": 1.6827466487884521,
+      "learning_rate": 6.021244309559939e-06,
+      "loss": 0.2458,
+      "step": 1485
+    },
+    {
+      "epoch": 0.42891887509683463,
+      "grad_norm": 4.625283241271973,
+      "learning_rate": 6.012139605462823e-06,
+      "loss": 0.2707,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4297836308934008,
+      "grad_norm": 2.733687400817871,
+      "learning_rate": 6.003034901365707e-06,
+      "loss": 0.2294,
+      "step": 1491
+    },
+    {
+      "epoch": 0.430648386689967,
+      "grad_norm": 1.38575279712677,
+      "learning_rate": 5.9939301972685896e-06,
+      "loss": 0.1992,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4315131424865332,
+      "grad_norm": 1.9684631824493408,
+      "learning_rate": 5.984825493171472e-06,
+      "loss": 0.2118,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4323778982830994,
+      "grad_norm": 3.40984845161438,
+      "learning_rate": 5.975720789074355e-06,
+      "loss": 0.2175,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4323778982830994,
+      "eval_loss": 0.2375136762857437,
+      "eval_mse": 0.23751368772797288,
+      "eval_runtime": 6.7445,
+      "eval_samples_per_second": 148.27,
+      "eval_steps_per_second": 18.534,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4332426540796656,
+      "grad_norm": 2.3788678646087646,
+      "learning_rate": 5.966616084977239e-06,
+      "loss": 0.2311,
+      "step": 1503
+    },
+    {
+      "epoch": 0.4341074098762318,
+      "grad_norm": 4.027227401733398,
+      "learning_rate": 5.957511380880122e-06,
+      "loss": 0.2456,
+      "step": 1506
+    },
+    {
+      "epoch": 0.434972165672798,
+      "grad_norm": 5.0818586349487305,
+      "learning_rate": 5.9484066767830055e-06,
+      "loss": 0.2631,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4358369214693642,
+      "grad_norm": 4.373122215270996,
+      "learning_rate": 5.939301972685888e-06,
+      "loss": 0.243,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4367016772659304,
+      "grad_norm": 3.31792950630188,
+      "learning_rate": 5.930197268588771e-06,
+      "loss": 0.2278,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4375664330624966,
+      "grad_norm": 1.9427984952926636,
+      "learning_rate": 5.921092564491655e-06,
+      "loss": 0.2578,
+      "step": 1518
+    },
+    {
+      "epoch": 0.43843118885906285,
+      "grad_norm": 2.5355935096740723,
+      "learning_rate": 5.911987860394538e-06,
+      "loss": 0.2125,
+      "step": 1521
+    },
+    {
+      "epoch": 0.43929594465562904,
+      "grad_norm": 5.661628723144531,
+      "learning_rate": 5.9028831562974205e-06,
+      "loss": 0.2591,
+      "step": 1524
+    },
+    {
+      "epoch": 0.44016070045219524,
+      "grad_norm": 2.133945941925049,
+      "learning_rate": 5.893778452200304e-06,
+      "loss": 0.2577,
+      "step": 1527
+    },
+    {
+      "epoch": 0.44102545624876144,
+      "grad_norm": 2.8841874599456787,
+      "learning_rate": 5.884673748103187e-06,
+      "loss": 0.2344,
+      "step": 1530
+    },
+    {
+      "epoch": 0.44189021204532763,
+      "grad_norm": 1.6562261581420898,
+      "learning_rate": 5.87556904400607e-06,
+      "loss": 0.2133,
+      "step": 1533
+    },
+    {
+      "epoch": 0.44275496784189383,
+      "grad_norm": 3.133864164352417,
+      "learning_rate": 5.866464339908954e-06,
+      "loss": 0.2154,
+      "step": 1536
+    },
+    {
+      "epoch": 0.44361972363846003,
+      "grad_norm": 1.9966986179351807,
+      "learning_rate": 5.8573596358118364e-06,
+      "loss": 0.1964,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4444844794350262,
+      "grad_norm": 1.9703294038772583,
+      "learning_rate": 5.848254931714719e-06,
+      "loss": 0.2238,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4453492352315924,
+      "grad_norm": 3.2984211444854736,
+      "learning_rate": 5.839150227617603e-06,
+      "loss": 0.2628,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4462139910281586,
+      "grad_norm": 3.6368560791015625,
+      "learning_rate": 5.830045523520486e-06,
+      "loss": 0.2567,
+      "step": 1548
+    },
+    {
+      "epoch": 0.4470787468247248,
+      "grad_norm": 2.366480827331543,
+      "learning_rate": 5.8209408194233695e-06,
+      "loss": 0.2493,
+      "step": 1551
+    },
+    {
+      "epoch": 0.447943502621291,
+      "grad_norm": 3.3239293098449707,
+      "learning_rate": 5.811836115326252e-06,
+      "loss": 0.2268,
+      "step": 1554
+    },
+    {
+      "epoch": 0.4488082584178572,
+      "grad_norm": 4.618416786193848,
+      "learning_rate": 5.802731411229135e-06,
+      "loss": 0.2456,
+      "step": 1557
+    },
+    {
+      "epoch": 0.4496730142144234,
+      "grad_norm": 2.826070785522461,
+      "learning_rate": 5.793626707132019e-06,
+      "loss": 0.2206,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4505377700109896,
+      "grad_norm": 4.012238025665283,
+      "learning_rate": 5.784522003034902e-06,
+      "loss": 0.2157,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4514025258075558,
+      "grad_norm": 2.0067975521087646,
+      "learning_rate": 5.775417298937785e-06,
+      "loss": 0.2489,
+      "step": 1566
+    },
+    {
+      "epoch": 0.452267281604122,
+      "grad_norm": 1.500907301902771,
+      "learning_rate": 5.766312594840668e-06,
+      "loss": 0.2348,
+      "step": 1569
+    },
+    {
+      "epoch": 0.4531320374006882,
+      "grad_norm": 1.8496533632278442,
+      "learning_rate": 5.757207890743551e-06,
+      "loss": 0.2019,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4539967931972544,
+      "grad_norm": 3.222740650177002,
+      "learning_rate": 5.748103186646434e-06,
+      "loss": 0.2587,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4548615489938206,
+      "grad_norm": 1.572925329208374,
+      "learning_rate": 5.738998482549318e-06,
+      "loss": 0.2265,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4557263047903868,
+      "grad_norm": 1.603624701499939,
+      "learning_rate": 5.729893778452201e-06,
+      "loss": 0.2433,
+      "step": 1581
+    },
+    {
+      "epoch": 0.456591060586953,
+      "grad_norm": 3.0979552268981934,
+      "learning_rate": 5.720789074355084e-06,
+      "loss": 0.221,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4574558163835192,
+      "grad_norm": 1.8691914081573486,
+      "learning_rate": 5.711684370257967e-06,
+      "loss": 0.2083,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4583205721800854,
+      "grad_norm": 1.6643935441970825,
+      "learning_rate": 5.70257966616085e-06,
+      "loss": 0.2173,
+      "step": 1590
+    },
+    {
+      "epoch": 0.45918532797665157,
+      "grad_norm": 3.626629590988159,
+      "learning_rate": 5.693474962063733e-06,
+      "loss": 0.2256,
+      "step": 1593
+    },
+    {
+      "epoch": 0.46005008377321777,
+      "grad_norm": 1.6160587072372437,
+      "learning_rate": 5.684370257966616e-06,
+      "loss": 0.2197,
+      "step": 1596
+    },
+    {
+      "epoch": 0.460914839569784,
+      "grad_norm": 3.206094264984131,
+      "learning_rate": 5.6752655538695e-06,
+      "loss": 0.2461,
+      "step": 1599
+    },
+    {
+      "epoch": 0.46120309150197275,
+      "eval_loss": 0.24926815927028656,
+      "eval_mse": 0.24926817585621028,
+      "eval_runtime": 6.5749,
+      "eval_samples_per_second": 152.094,
+      "eval_steps_per_second": 19.012,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4617795953663502,
+      "grad_norm": 4.698122024536133,
+      "learning_rate": 5.666160849772383e-06,
+      "loss": 0.2546,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4626443511629164,
+      "grad_norm": 1.5181471109390259,
+      "learning_rate": 5.657056145675266e-06,
+      "loss": 0.2167,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4635091069594826,
+      "grad_norm": 7.337409496307373,
+      "learning_rate": 5.647951441578149e-06,
+      "loss": 0.2435,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4643738627560488,
+      "grad_norm": 1.6400500535964966,
+      "learning_rate": 5.6388467374810314e-06,
+      "loss": 0.2336,
+      "step": 1611
+    },
+    {
+      "epoch": 0.465238618552615,
+      "grad_norm": 1.7364429235458374,
+      "learning_rate": 5.629742033383916e-06,
+      "loss": 0.2397,
+      "step": 1614
+    },
+    {
+      "epoch": 0.4661033743491812,
+      "grad_norm": 2.7070679664611816,
+      "learning_rate": 5.620637329286799e-06,
+      "loss": 0.2389,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4669681301457474,
+      "grad_norm": 1.28640878200531,
+      "learning_rate": 5.611532625189682e-06,
+      "loss": 0.202,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4678328859423136,
+      "grad_norm": 2.6867973804473877,
+      "learning_rate": 5.6024279210925645e-06,
+      "loss": 0.2501,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4686976417388798,
+      "grad_norm": 2.1441292762756348,
+      "learning_rate": 5.593323216995447e-06,
+      "loss": 0.2364,
+      "step": 1626
+    },
+    {
+      "epoch": 0.469562397535446,
+      "grad_norm": 1.6109544038772583,
+      "learning_rate": 5.584218512898332e-06,
+      "loss": 0.244,
+      "step": 1629
+    },
+    {
+      "epoch": 0.4704271533320122,
+      "grad_norm": 2.0842268466949463,
+      "learning_rate": 5.575113808801215e-06,
+      "loss": 0.2434,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4712919091285784,
+      "grad_norm": 1.4527335166931152,
+      "learning_rate": 5.5660091047040976e-06,
+      "loss": 0.2113,
+      "step": 1635
+    },
+    {
+      "epoch": 0.4721566649251446,
+      "grad_norm": 2.0434927940368652,
+      "learning_rate": 5.55690440060698e-06,
+      "loss": 0.2404,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4730214207217108,
+      "grad_norm": 3.0256736278533936,
+      "learning_rate": 5.547799696509863e-06,
+      "loss": 0.2652,
+      "step": 1641
+    },
+    {
+      "epoch": 0.47388617651827697,
+      "grad_norm": 2.3856120109558105,
+      "learning_rate": 5.538694992412748e-06,
+      "loss": 0.2156,
+      "step": 1644
+    },
+    {
+      "epoch": 0.47475093231484317,
+      "grad_norm": 2.0948779582977295,
+      "learning_rate": 5.529590288315631e-06,
+      "loss": 0.2245,
+      "step": 1647
+    },
+    {
+      "epoch": 0.47561568811140936,
+      "grad_norm": 1.7975860834121704,
+      "learning_rate": 5.5204855842185135e-06,
+      "loss": 0.2542,
+      "step": 1650
+    },
+    {
+      "epoch": 0.47648044390797556,
+      "grad_norm": 3.510812997817993,
+      "learning_rate": 5.511380880121396e-06,
+      "loss": 0.2288,
+      "step": 1653
+    },
+    {
+      "epoch": 0.47734519970454176,
+      "grad_norm": 3.4858286380767822,
+      "learning_rate": 5.502276176024279e-06,
+      "loss": 0.2497,
+      "step": 1656
+    },
+    {
+      "epoch": 0.47820995550110795,
+      "grad_norm": 2.605661630630493,
+      "learning_rate": 5.493171471927162e-06,
+      "loss": 0.2062,
+      "step": 1659
+    },
+    {
+      "epoch": 0.47907471129767415,
+      "grad_norm": 2.162203788757324,
+      "learning_rate": 5.4840667678300465e-06,
+      "loss": 0.2167,
+      "step": 1662
+    },
+    {
+      "epoch": 0.47993946709424035,
+      "grad_norm": 4.7574262619018555,
+      "learning_rate": 5.474962063732929e-06,
+      "loss": 0.2088,
+      "step": 1665
+    },
+    {
+      "epoch": 0.48080422289080654,
+      "grad_norm": 2.1833505630493164,
+      "learning_rate": 5.465857359635812e-06,
+      "loss": 0.2622,
+      "step": 1668
+    },
+    {
+      "epoch": 0.48166897868737274,
+      "grad_norm": 1.645045280456543,
+      "learning_rate": 5.456752655538695e-06,
+      "loss": 0.2164,
+      "step": 1671
+    },
+    {
+      "epoch": 0.48253373448393894,
+      "grad_norm": 2.270944356918335,
+      "learning_rate": 5.447647951441578e-06,
+      "loss": 0.2441,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4833984902805052,
+      "grad_norm": 2.0133309364318848,
+      "learning_rate": 5.4385432473444624e-06,
+      "loss": 0.2032,
+      "step": 1677
+    },
+    {
+      "epoch": 0.4842632460770714,
+      "grad_norm": 2.5585310459136963,
+      "learning_rate": 5.429438543247345e-06,
+      "loss": 0.2385,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4851280018736376,
+      "grad_norm": 1.7377713918685913,
+      "learning_rate": 5.420333839150228e-06,
+      "loss": 0.216,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4859927576702038,
+      "grad_norm": 1.8322491645812988,
+      "learning_rate": 5.411229135053111e-06,
+      "loss": 0.2102,
+      "step": 1686
+    },
+    {
+      "epoch": 0.48685751346677,
+      "grad_norm": 1.8956815004348755,
+      "learning_rate": 5.402124430955994e-06,
+      "loss": 0.2561,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4877222692633362,
+      "grad_norm": 1.5535213947296143,
+      "learning_rate": 5.393019726858878e-06,
+      "loss": 0.2099,
+      "step": 1692
+    },
+    {
+      "epoch": 0.48858702505990237,
+      "grad_norm": 2.924278974533081,
+      "learning_rate": 5.383915022761761e-06,
+      "loss": 0.2662,
+      "step": 1695
+    },
+    {
+      "epoch": 0.48945178085646857,
+      "grad_norm": 2.1653637886047363,
+      "learning_rate": 5.374810318664644e-06,
+      "loss": 0.2419,
+      "step": 1698
+    },
+    {
+      "epoch": 0.49002828472084603,
+      "eval_loss": 0.22344937920570374,
+      "eval_mse": 0.22344938813522458,
+      "eval_runtime": 6.5992,
+      "eval_samples_per_second": 151.533,
+      "eval_steps_per_second": 18.942,
+      "step": 1700
+    },
+    {
+      "epoch": 0.49031653665303476,
+      "grad_norm": 2.360328197479248,
+      "learning_rate": 5.365705614567527e-06,
+      "loss": 0.2661,
+      "step": 1701
+    },
+    {
+      "epoch": 0.49118129244960096,
+      "grad_norm": 2.328495502471924,
+      "learning_rate": 5.35660091047041e-06,
+      "loss": 0.2347,
+      "step": 1704
+    },
+    {
+      "epoch": 0.49204604824616716,
+      "grad_norm": 1.6670514345169067,
+      "learning_rate": 5.347496206373293e-06,
+      "loss": 0.2269,
+      "step": 1707
+    },
+    {
+      "epoch": 0.49291080404273335,
+      "grad_norm": 2.426805257797241,
+      "learning_rate": 5.338391502276177e-06,
+      "loss": 0.1946,
+      "step": 1710
+    },
+    {
+      "epoch": 0.49377555983929955,
+      "grad_norm": 1.7583879232406616,
+      "learning_rate": 5.32928679817906e-06,
+      "loss": 0.2235,
+      "step": 1713
+    },
+    {
+      "epoch": 0.49464031563586575,
+      "grad_norm": 1.7235326766967773,
+      "learning_rate": 5.320182094081943e-06,
+      "loss": 0.2395,
+      "step": 1716
+    },
+    {
+      "epoch": 0.49550507143243194,
+      "grad_norm": 1.4216803312301636,
+      "learning_rate": 5.311077389984826e-06,
+      "loss": 0.2634,
+      "step": 1719
+    },
+    {
+      "epoch": 0.49636982722899814,
+      "grad_norm": 1.2892686128616333,
+      "learning_rate": 5.301972685887709e-06,
+      "loss": 0.2083,
+      "step": 1722
+    },
+    {
+      "epoch": 0.49723458302556434,
+      "grad_norm": 2.5210540294647217,
+      "learning_rate": 5.292867981790593e-06,
+      "loss": 0.2017,
+      "step": 1725
+    },
+    {
+      "epoch": 0.49809933882213053,
+      "grad_norm": 5.790046691894531,
+      "learning_rate": 5.283763277693476e-06,
+      "loss": 0.2306,
+      "step": 1728
+    },
+    {
+      "epoch": 0.49896409461869673,
+      "grad_norm": 1.4158023595809937,
+      "learning_rate": 5.274658573596359e-06,
+      "loss": 0.266,
+      "step": 1731
+    },
+    {
+      "epoch": 0.4998288504152629,
+      "grad_norm": 3.0490024089813232,
+      "learning_rate": 5.2655538694992415e-06,
+      "loss": 0.2506,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5006936062118291,
+      "grad_norm": 2.33208966255188,
+      "learning_rate": 5.256449165402124e-06,
+      "loss": 0.2088,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5015583620083953,
+      "grad_norm": 1.8087997436523438,
+      "learning_rate": 5.247344461305008e-06,
+      "loss": 0.2095,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5024231178049615,
+      "grad_norm": 1.5517979860305786,
+      "learning_rate": 5.238239757207892e-06,
+      "loss": 0.2342,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5032878736015277,
+      "grad_norm": 1.841036319732666,
+      "learning_rate": 5.229135053110775e-06,
+      "loss": 0.2385,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5041526293980939,
+      "grad_norm": 1.8034095764160156,
+      "learning_rate": 5.2200303490136574e-06,
+      "loss": 0.2537,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5050173851946601,
+      "grad_norm": 3.617159366607666,
+      "learning_rate": 5.21092564491654e-06,
+      "loss": 0.2378,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5058821409912263,
+      "grad_norm": 2.903215169906616,
+      "learning_rate": 5.201820940819424e-06,
+      "loss": 0.2356,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5067468967877925,
+      "grad_norm": 2.084693193435669,
+      "learning_rate": 5.192716236722307e-06,
+      "loss": 0.2631,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5076116525843587,
+      "grad_norm": 1.8994488716125488,
+      "learning_rate": 5.1836115326251905e-06,
+      "loss": 0.2742,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5084764083809249,
+      "grad_norm": 2.651257276535034,
+      "learning_rate": 5.174506828528073e-06,
+      "loss": 0.2332,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5093411641774911,
+      "grad_norm": 4.182311534881592,
+      "learning_rate": 5.165402124430956e-06,
+      "loss": 0.2467,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5102059199740573,
+      "grad_norm": 27.990720748901367,
+      "learning_rate": 5.15629742033384e-06,
+      "loss": 0.2135,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5110706757706235,
+      "grad_norm": 1.942474126815796,
+      "learning_rate": 5.147192716236723e-06,
+      "loss": 0.2443,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5119354315671897,
+      "grad_norm": 2.768105983734131,
+      "learning_rate": 5.1380880121396055e-06,
+      "loss": 0.2566,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5128001873637559,
+      "grad_norm": 2.423797607421875,
+      "learning_rate": 5.128983308042489e-06,
+      "loss": 0.2667,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5136649431603221,
+      "grad_norm": 2.395047426223755,
+      "learning_rate": 5.119878603945372e-06,
+      "loss": 0.2495,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5145296989568883,
+      "grad_norm": 2.577787160873413,
+      "learning_rate": 5.110773899848256e-06,
+      "loss": 0.2454,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5153944547534545,
+      "grad_norm": 3.106776714324951,
+      "learning_rate": 5.101669195751139e-06,
+      "loss": 0.2162,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5162592105500207,
+      "grad_norm": 1.5912446975708008,
+      "learning_rate": 5.0925644916540215e-06,
+      "loss": 0.2201,
+      "step": 1791
+    },
+    {
+      "epoch": 0.517123966346587,
+      "grad_norm": 2.427795171737671,
+      "learning_rate": 5.083459787556905e-06,
+      "loss": 0.236,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5179887221431532,
+      "grad_norm": 2.5363399982452393,
+      "learning_rate": 5.074355083459788e-06,
+      "loss": 0.2551,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5188534779397194,
+      "grad_norm": 1.8077950477600098,
+      "learning_rate": 5.065250379362671e-06,
+      "loss": 0.2411,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5188534779397194,
+      "eval_loss": 0.21374772489070892,
+      "eval_mse": 0.21374773593991994,
+      "eval_runtime": 6.6116,
+      "eval_samples_per_second": 151.249,
+      "eval_steps_per_second": 18.906,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5197182337362856,
+      "grad_norm": 5.243107795715332,
+      "learning_rate": 5.0561456752655545e-06,
+      "loss": 0.243,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5205829895328518,
+      "grad_norm": 2.7711331844329834,
+      "learning_rate": 5.047040971168437e-06,
+      "loss": 0.2037,
+      "step": 1806
+    },
+    {
+      "epoch": 0.521447745329418,
+      "grad_norm": 2.8418238162994385,
+      "learning_rate": 5.03793626707132e-06,
+      "loss": 0.2331,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5223125011259842,
+      "grad_norm": 3.0842299461364746,
+      "learning_rate": 5.028831562974204e-06,
+      "loss": 0.2422,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5231772569225503,
+      "grad_norm": 2.5956835746765137,
+      "learning_rate": 5.019726858877087e-06,
+      "loss": 0.2057,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5240420127191165,
+      "grad_norm": 1.3121715784072876,
+      "learning_rate": 5.0106221547799704e-06,
+      "loss": 0.2387,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5249067685156827,
+      "grad_norm": 2.2341432571411133,
+      "learning_rate": 5.001517450682853e-06,
+      "loss": 0.2611,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5257715243122489,
+      "grad_norm": 2.0494587421417236,
+      "learning_rate": 4.992412746585736e-06,
+      "loss": 0.2238,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5266362801088151,
+      "grad_norm": 2.2597897052764893,
+      "learning_rate": 4.983308042488619e-06,
+      "loss": 0.2322,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5275010359053813,
+      "grad_norm": 3.5993051528930664,
+      "learning_rate": 4.974203338391503e-06,
+      "loss": 0.234,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5283657917019475,
+      "grad_norm": 3.769505262374878,
+      "learning_rate": 4.9650986342943855e-06,
+      "loss": 0.2454,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5292305474985137,
+      "grad_norm": 1.4726693630218506,
+      "learning_rate": 4.955993930197269e-06,
+      "loss": 0.2261,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5300953032950799,
+      "grad_norm": 2.2910659313201904,
+      "learning_rate": 4.946889226100152e-06,
+      "loss": 0.2551,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5309600590916461,
+      "grad_norm": 1.8401825428009033,
+      "learning_rate": 4.937784522003035e-06,
+      "loss": 0.1891,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5318248148882123,
+      "grad_norm": 2.0589776039123535,
+      "learning_rate": 4.9286798179059185e-06,
+      "loss": 0.2475,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5326895706847785,
+      "grad_norm": 1.8461434841156006,
+      "learning_rate": 4.919575113808801e-06,
+      "loss": 0.2516,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5335543264813447,
+      "grad_norm": 1.8950605392456055,
+      "learning_rate": 4.910470409711684e-06,
+      "loss": 0.246,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5344190822779109,
+      "grad_norm": 1.7544567584991455,
+      "learning_rate": 4.901365705614568e-06,
+      "loss": 0.2159,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5352838380744771,
+      "grad_norm": 2.04953932762146,
+      "learning_rate": 4.892261001517451e-06,
+      "loss": 0.2745,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5361485938710433,
+      "grad_norm": 2.173112154006958,
+      "learning_rate": 4.8831562974203345e-06,
+      "loss": 0.2222,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5370133496676095,
+      "grad_norm": 1.4713711738586426,
+      "learning_rate": 4.874051593323217e-06,
+      "loss": 0.2182,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5378781054641757,
+      "grad_norm": 2.421405792236328,
+      "learning_rate": 4.8649468892261e-06,
+      "loss": 0.2472,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5387428612607419,
+      "grad_norm": 2.6306304931640625,
+      "learning_rate": 4.855842185128984e-06,
+      "loss": 0.2148,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5396076170573081,
+      "grad_norm": 3.4957375526428223,
+      "learning_rate": 4.846737481031867e-06,
+      "loss": 0.2321,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5404723728538743,
+      "grad_norm": 4.008154392242432,
+      "learning_rate": 4.8376327769347495e-06,
+      "loss": 0.216,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5413371286504405,
+      "grad_norm": 2.417433977127075,
+      "learning_rate": 4.828528072837633e-06,
+      "loss": 0.2261,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5422018844470067,
+      "grad_norm": 1.7908028364181519,
+      "learning_rate": 4.819423368740516e-06,
+      "loss": 0.2445,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5430666402435729,
+      "grad_norm": 1.3756656646728516,
+      "learning_rate": 4.8103186646434e-06,
+      "loss": 0.201,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5439313960401391,
+      "grad_norm": 1.686787724494934,
+      "learning_rate": 4.801213960546283e-06,
+      "loss": 0.217,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5447961518367053,
+      "grad_norm": 3.6207942962646484,
+      "learning_rate": 4.792109256449165e-06,
+      "loss": 0.2746,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5456609076332715,
+      "grad_norm": 3.838956117630005,
+      "learning_rate": 4.783004552352049e-06,
+      "loss": 0.2313,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5465256634298377,
+      "grad_norm": 2.059926748275757,
+      "learning_rate": 4.773899848254932e-06,
+      "loss": 0.2302,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5473904192264039,
+      "grad_norm": 1.9524738788604736,
+      "learning_rate": 4.764795144157816e-06,
+      "loss": 0.2473,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5476786711585926,
+      "eval_loss": 0.21403329074382782,
+      "eval_mse": 0.2140333094932139,
+      "eval_runtime": 6.5394,
+      "eval_samples_per_second": 152.918,
+      "eval_steps_per_second": 19.115,
+      "step": 1900
+    },
+    {
+      "epoch": 0.54825517502297,
+      "grad_norm": 1.8144432306289673,
+      "learning_rate": 4.7556904400606985e-06,
+      "loss": 0.2082,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5491199308195363,
+      "grad_norm": 2.4071717262268066,
+      "learning_rate": 4.746585735963581e-06,
+      "loss": 0.2364,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5499846866161024,
+      "grad_norm": 1.7162179946899414,
+      "learning_rate": 4.737481031866465e-06,
+      "loss": 0.2119,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5508494424126686,
+      "grad_norm": 2.368528366088867,
+      "learning_rate": 4.728376327769348e-06,
+      "loss": 0.2314,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5517141982092348,
+      "grad_norm": 3.422670602798462,
+      "learning_rate": 4.719271623672231e-06,
+      "loss": 0.2355,
+      "step": 1914
+    },
+    {
+      "epoch": 0.552578954005801,
+      "grad_norm": 2.324976682662964,
+      "learning_rate": 4.710166919575114e-06,
+      "loss": 0.2483,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5534437098023672,
+      "grad_norm": 2.3686418533325195,
+      "learning_rate": 4.701062215477997e-06,
+      "loss": 0.204,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5543084655989334,
+      "grad_norm": 2.1361286640167236,
+      "learning_rate": 4.691957511380881e-06,
+      "loss": 0.195,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5551732213954996,
+      "grad_norm": 1.5527316331863403,
+      "learning_rate": 4.682852807283764e-06,
+      "loss": 0.2232,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5560379771920658,
+      "grad_norm": 3.792592763900757,
+      "learning_rate": 4.673748103186647e-06,
+      "loss": 0.2157,
+      "step": 1929
+    },
+    {
+      "epoch": 0.556902732988632,
+      "grad_norm": 1.8878562450408936,
+      "learning_rate": 4.66464339908953e-06,
+      "loss": 0.21,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5577674887851982,
+      "grad_norm": 1.686164140701294,
+      "learning_rate": 4.655538694992413e-06,
+      "loss": 0.2177,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5586322445817644,
+      "grad_norm": 2.310054302215576,
+      "learning_rate": 4.646433990895296e-06,
+      "loss": 0.2039,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5594970003783306,
+      "grad_norm": 1.8163293600082397,
+      "learning_rate": 4.63732928679818e-06,
+      "loss": 0.2117,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5603617561748968,
+      "grad_norm": 1.608209490776062,
+      "learning_rate": 4.6282245827010625e-06,
+      "loss": 0.2361,
+      "step": 1944
+    },
+    {
+      "epoch": 0.561226511971463,
+      "grad_norm": 1.4159106016159058,
+      "learning_rate": 4.619119878603946e-06,
+      "loss": 0.2042,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5620912677680293,
+      "grad_norm": 2.920888900756836,
+      "learning_rate": 4.610015174506829e-06,
+      "loss": 0.2715,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5629560235645955,
+      "grad_norm": 1.1805585622787476,
+      "learning_rate": 4.600910470409712e-06,
+      "loss": 0.2329,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5638207793611617,
+      "grad_norm": 3.1957271099090576,
+      "learning_rate": 4.591805766312596e-06,
+      "loss": 0.253,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5646855351577279,
+      "grad_norm": 2.3281495571136475,
+      "learning_rate": 4.582701062215478e-06,
+      "loss": 0.2508,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5655502909542941,
+      "grad_norm": 1.9826480150222778,
+      "learning_rate": 4.573596358118362e-06,
+      "loss": 0.2107,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5664150467508603,
+      "grad_norm": 5.98641300201416,
+      "learning_rate": 4.564491654021245e-06,
+      "loss": 0.2343,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5672798025474265,
+      "grad_norm": 4.082788944244385,
+      "learning_rate": 4.555386949924128e-06,
+      "loss": 0.2493,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5681445583439927,
+      "grad_norm": 1.6683293581008911,
+      "learning_rate": 4.5462822458270115e-06,
+      "loss": 0.1892,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5690093141405589,
+      "grad_norm": 4.348227024078369,
+      "learning_rate": 4.537177541729894e-06,
+      "loss": 0.2277,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5698740699371251,
+      "grad_norm": 1.2774113416671753,
+      "learning_rate": 4.528072837632777e-06,
+      "loss": 0.2077,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5707388257336913,
+      "grad_norm": 1.5915591716766357,
+      "learning_rate": 4.518968133535661e-06,
+      "loss": 0.2284,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5716035815302575,
+      "grad_norm": 1.3817569017410278,
+      "learning_rate": 4.509863429438544e-06,
+      "loss": 0.2159,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5724683373268237,
+      "grad_norm": 1.3606081008911133,
+      "learning_rate": 4.500758725341427e-06,
+      "loss": 0.2055,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5733330931233899,
+      "grad_norm": 2.6602344512939453,
+      "learning_rate": 4.49165402124431e-06,
+      "loss": 0.2611,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5741978489199561,
+      "grad_norm": 1.796570897102356,
+      "learning_rate": 4.482549317147193e-06,
+      "loss": 0.1925,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5750626047165223,
+      "grad_norm": 2.4218592643737793,
+      "learning_rate": 4.473444613050077e-06,
+      "loss": 0.1992,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5759273605130885,
+      "grad_norm": 1.9690345525741577,
+      "learning_rate": 4.46433990895296e-06,
+      "loss": 0.237,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5765038643774659,
+      "eval_loss": 0.2176571637392044,
+      "eval_mse": 0.2176571699755732,
+      "eval_runtime": 6.5589,
+      "eval_samples_per_second": 152.465,
+      "eval_steps_per_second": 19.058,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5767921163096547,
+      "grad_norm": 2.168555498123169,
+      "learning_rate": 4.4552352048558425e-06,
+      "loss": 0.2009,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5776568721062209,
+      "grad_norm": 3.1508069038391113,
+      "learning_rate": 4.446130500758726e-06,
+      "loss": 0.2316,
+      "step": 2004
+    },
+    {
+      "epoch": 0.578521627902787,
+      "grad_norm": 1.686283826828003,
+      "learning_rate": 4.437025796661609e-06,
+      "loss": 0.2313,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5793863836993532,
+      "grad_norm": 1.3212209939956665,
+      "learning_rate": 4.427921092564492e-06,
+      "loss": 0.2076,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5802511394959194,
+      "grad_norm": 3.2884185314178467,
+      "learning_rate": 4.4188163884673755e-06,
+      "loss": 0.2345,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5811158952924856,
+      "grad_norm": 1.9892568588256836,
+      "learning_rate": 4.409711684370258e-06,
+      "loss": 0.2428,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5819806510890518,
+      "grad_norm": 1.5294564962387085,
+      "learning_rate": 4.400606980273141e-06,
+      "loss": 0.2252,
+      "step": 2019
+    },
+    {
+      "epoch": 0.582845406885618,
+      "grad_norm": 2.2944376468658447,
+      "learning_rate": 4.391502276176025e-06,
+      "loss": 0.2429,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5837101626821842,
+      "grad_norm": 1.584145426750183,
+      "learning_rate": 4.382397572078908e-06,
+      "loss": 0.2548,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5845749184787504,
+      "grad_norm": 1.604771375656128,
+      "learning_rate": 4.3732928679817906e-06,
+      "loss": 0.2211,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5854396742753166,
+      "grad_norm": 1.7536587715148926,
+      "learning_rate": 4.364188163884674e-06,
+      "loss": 0.2165,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5863044300718828,
+      "grad_norm": 1.6281161308288574,
+      "learning_rate": 4.355083459787557e-06,
+      "loss": 0.2609,
+      "step": 2034
+    },
+    {
+      "epoch": 0.587169185868449,
+      "grad_norm": 1.77180016040802,
+      "learning_rate": 4.34597875569044e-06,
+      "loss": 0.2333,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5880339416650152,
+      "grad_norm": 3.2408559322357178,
+      "learning_rate": 4.336874051593324e-06,
+      "loss": 0.2454,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5888986974615814,
+      "grad_norm": 3.4444427490234375,
+      "learning_rate": 4.3277693474962065e-06,
+      "loss": 0.2549,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5897634532581476,
+      "grad_norm": 4.399412155151367,
+      "learning_rate": 4.31866464339909e-06,
+      "loss": 0.2258,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5906282090547138,
+      "grad_norm": 1.646681785583496,
+      "learning_rate": 4.309559939301973e-06,
+      "loss": 0.1858,
+      "step": 2049
+    },
+    {
+      "epoch": 0.59149296485128,
+      "grad_norm": 2.786576986312866,
+      "learning_rate": 4.300455235204856e-06,
+      "loss": 0.2085,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5923577206478462,
+      "grad_norm": 3.0838379859924316,
+      "learning_rate": 4.2913505311077395e-06,
+      "loss": 0.2428,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5932224764444124,
+      "grad_norm": 1.518548846244812,
+      "learning_rate": 4.282245827010622e-06,
+      "loss": 0.2159,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5940872322409786,
+      "grad_norm": 2.2088887691497803,
+      "learning_rate": 4.273141122913505e-06,
+      "loss": 0.2107,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5949519880375448,
+      "grad_norm": 4.359911918640137,
+      "learning_rate": 4.264036418816389e-06,
+      "loss": 0.2461,
+      "step": 2064
+    },
+    {
+      "epoch": 0.595816743834111,
+      "grad_norm": 1.796669840812683,
+      "learning_rate": 4.254931714719272e-06,
+      "loss": 0.1836,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5966814996306772,
+      "grad_norm": 2.0835959911346436,
+      "learning_rate": 4.245827010622155e-06,
+      "loss": 0.2192,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5975462554272434,
+      "grad_norm": 2.772815227508545,
+      "learning_rate": 4.236722306525038e-06,
+      "loss": 0.1974,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5984110112238096,
+      "grad_norm": 1.8529541492462158,
+      "learning_rate": 4.227617602427921e-06,
+      "loss": 0.1946,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5992757670203758,
+      "grad_norm": 2.0415849685668945,
+      "learning_rate": 4.218512898330804e-06,
+      "loss": 0.2182,
+      "step": 2079
+    },
+    {
+      "epoch": 0.600140522816942,
+      "grad_norm": 2.6295053958892822,
+      "learning_rate": 4.209408194233688e-06,
+      "loss": 0.2438,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6010052786135082,
+      "grad_norm": 1.9082621335983276,
+      "learning_rate": 4.2003034901365705e-06,
+      "loss": 0.1834,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6018700344100744,
+      "grad_norm": 1.692436933517456,
+      "learning_rate": 4.191198786039454e-06,
+      "loss": 0.2104,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6027347902066406,
+      "grad_norm": 2.0836997032165527,
+      "learning_rate": 4.182094081942337e-06,
+      "loss": 0.2323,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6035995460032068,
+      "grad_norm": 1.354612946510315,
+      "learning_rate": 4.17298937784522e-06,
+      "loss": 0.1975,
+      "step": 2094
+    },
+    {
+      "epoch": 0.604464301799773,
+      "grad_norm": 3.683278799057007,
+      "learning_rate": 4.1638846737481036e-06,
+      "loss": 0.2256,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6053290575963391,
+      "grad_norm": 2.232513189315796,
+      "learning_rate": 4.154779969650986e-06,
+      "loss": 0.1972,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6053290575963391,
+      "eval_loss": 0.21862919628620148,
+      "eval_mse": 0.218629194105044,
+      "eval_runtime": 6.7266,
+      "eval_samples_per_second": 148.664,
+      "eval_steps_per_second": 18.583,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6061938133929053,
+      "grad_norm": 1.5396499633789062,
+      "learning_rate": 4.145675265553869e-06,
+      "loss": 0.2151,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6070585691894717,
+      "grad_norm": 1.482253909111023,
+      "learning_rate": 4.136570561456753e-06,
+      "loss": 0.2222,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6079233249860378,
+      "grad_norm": 4.880987644195557,
+      "learning_rate": 4.127465857359636e-06,
+      "loss": 0.2535,
+      "step": 2109
+    },
+    {
+      "epoch": 0.608788080782604,
+      "grad_norm": 2.1557230949401855,
+      "learning_rate": 4.1183611532625195e-06,
+      "loss": 0.2213,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6096528365791702,
+      "grad_norm": 2.417856454849243,
+      "learning_rate": 4.109256449165402e-06,
+      "loss": 0.2481,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6105175923757364,
+      "grad_norm": 2.211514949798584,
+      "learning_rate": 4.100151745068285e-06,
+      "loss": 0.2369,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6113823481723026,
+      "grad_norm": 2.2844600677490234,
+      "learning_rate": 4.091047040971169e-06,
+      "loss": 0.2174,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6122471039688688,
+      "grad_norm": 2.7534289360046387,
+      "learning_rate": 4.081942336874052e-06,
+      "loss": 0.22,
+      "step": 2124
+    },
+    {
+      "epoch": 0.613111859765435,
+      "grad_norm": 1.5547044277191162,
+      "learning_rate": 4.072837632776935e-06,
+      "loss": 0.227,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6139766155620012,
+      "grad_norm": 1.6965092420578003,
+      "learning_rate": 4.063732928679818e-06,
+      "loss": 0.219,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6148413713585674,
+      "grad_norm": 1.8000000715255737,
+      "learning_rate": 4.054628224582701e-06,
+      "loss": 0.2303,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6157061271551336,
+      "grad_norm": 1.5764316320419312,
+      "learning_rate": 4.045523520485585e-06,
+      "loss": 0.191,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6165708829516998,
+      "grad_norm": 2.0041658878326416,
+      "learning_rate": 4.036418816388468e-06,
+      "loss": 0.2025,
+      "step": 2139
+    },
+    {
+      "epoch": 0.617435638748266,
+      "grad_norm": 1.9013463258743286,
+      "learning_rate": 4.0273141122913504e-06,
+      "loss": 0.225,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6183003945448322,
+      "grad_norm": 2.1815786361694336,
+      "learning_rate": 4.018209408194234e-06,
+      "loss": 0.2409,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6191651503413984,
+      "grad_norm": 1.6740418672561646,
+      "learning_rate": 4.009104704097117e-06,
+      "loss": 0.2143,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6200299061379646,
+      "grad_norm": 2.1056814193725586,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.2365,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6208946619345308,
+      "grad_norm": 2.629563808441162,
+      "learning_rate": 3.9908952959028835e-06,
+      "loss": 0.1932,
+      "step": 2154
+    },
+    {
+      "epoch": 0.621759417731097,
+      "grad_norm": 1.7547650337219238,
+      "learning_rate": 3.981790591805766e-06,
+      "loss": 0.246,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6226241735276632,
+      "grad_norm": 2.2451794147491455,
+      "learning_rate": 3.97268588770865e-06,
+      "loss": 0.2405,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6234889293242294,
+      "grad_norm": 2.8820624351501465,
+      "learning_rate": 3.963581183611533e-06,
+      "loss": 0.253,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6243536851207956,
+      "grad_norm": 2.9832215309143066,
+      "learning_rate": 3.9544764795144166e-06,
+      "loss": 0.223,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6252184409173618,
+      "grad_norm": 2.911879539489746,
+      "learning_rate": 3.945371775417299e-06,
+      "loss": 0.2177,
+      "step": 2169
+    },
+    {
+      "epoch": 0.626083196713928,
+      "grad_norm": 2.266767740249634,
+      "learning_rate": 3.936267071320182e-06,
+      "loss": 0.2288,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6269479525104942,
+      "grad_norm": 1.401633858680725,
+      "learning_rate": 3.927162367223066e-06,
+      "loss": 0.2556,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6278127083070604,
+      "grad_norm": 2.7354822158813477,
+      "learning_rate": 3.918057663125949e-06,
+      "loss": 0.2439,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6286774641036266,
+      "grad_norm": 1.6652506589889526,
+      "learning_rate": 3.908952959028832e-06,
+      "loss": 0.2041,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6295422199001928,
+      "grad_norm": 3.3072733879089355,
+      "learning_rate": 3.899848254931715e-06,
+      "loss": 0.2207,
+      "step": 2184
+    },
+    {
+      "epoch": 0.630406975696759,
+      "grad_norm": 2.254608392715454,
+      "learning_rate": 3.890743550834598e-06,
+      "loss": 0.2352,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6312717314933252,
+      "grad_norm": 1.765981674194336,
+      "learning_rate": 3.881638846737482e-06,
+      "loss": 0.2923,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6321364872898914,
+      "grad_norm": 1.7792342901229858,
+      "learning_rate": 3.872534142640365e-06,
+      "loss": 0.2428,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6330012430864576,
+      "grad_norm": 5.084781646728516,
+      "learning_rate": 3.8634294385432475e-06,
+      "loss": 0.2616,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6338659988830238,
+      "grad_norm": 2.0845305919647217,
+      "learning_rate": 3.854324734446131e-06,
+      "loss": 0.2556,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6341542508152125,
+      "eval_loss": 0.2416454404592514,
+      "eval_mse": 0.2416454482518602,
+      "eval_runtime": 6.5442,
+      "eval_samples_per_second": 152.808,
+      "eval_steps_per_second": 19.101,
+      "step": 2200
+    },
+    {
+      "epoch": 0.63473075467959,
+      "grad_norm": 3.8279173374176025,
+      "learning_rate": 3.845220030349014e-06,
+      "loss": 0.2245,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6355955104761561,
+      "grad_norm": 3.564417839050293,
+      "learning_rate": 3.836115326251897e-06,
+      "loss": 0.2233,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6364602662727223,
+      "grad_norm": 2.4322614669799805,
+      "learning_rate": 3.827010622154781e-06,
+      "loss": 0.2227,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6373250220692885,
+      "grad_norm": 1.6439640522003174,
+      "learning_rate": 3.8179059180576634e-06,
+      "loss": 0.2135,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6381897778658547,
+      "grad_norm": 2.605598211288452,
+      "learning_rate": 3.8088012139605467e-06,
+      "loss": 0.2203,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6390545336624209,
+      "grad_norm": 1.939488410949707,
+      "learning_rate": 3.7996965098634296e-06,
+      "loss": 0.2377,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6399192894589871,
+      "grad_norm": 1.851778507232666,
+      "learning_rate": 3.790591805766313e-06,
+      "loss": 0.2199,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6407840452555533,
+      "grad_norm": 2.077923059463501,
+      "learning_rate": 3.781487101669196e-06,
+      "loss": 0.223,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6416488010521195,
+      "grad_norm": 5.192010402679443,
+      "learning_rate": 3.772382397572079e-06,
+      "loss": 0.2412,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6425135568486857,
+      "grad_norm": 1.6057194471359253,
+      "learning_rate": 3.7632776934749626e-06,
+      "loss": 0.2354,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6433783126452519,
+      "grad_norm": 3.0130653381347656,
+      "learning_rate": 3.7541729893778455e-06,
+      "loss": 0.2307,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6442430684418181,
+      "grad_norm": 2.835080623626709,
+      "learning_rate": 3.7450682852807287e-06,
+      "loss": 0.1929,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6451078242383843,
+      "grad_norm": 1.800140619277954,
+      "learning_rate": 3.735963581183612e-06,
+      "loss": 0.2158,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6459725800349505,
+      "grad_norm": 1.8859021663665771,
+      "learning_rate": 3.726858877086495e-06,
+      "loss": 0.2469,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6468373358315167,
+      "grad_norm": 1.8524531126022339,
+      "learning_rate": 3.717754172989378e-06,
+      "loss": 0.2257,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6477020916280829,
+      "grad_norm": 2.759021520614624,
+      "learning_rate": 3.7086494688922614e-06,
+      "loss": 0.2187,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6485668474246491,
+      "grad_norm": 1.885272741317749,
+      "learning_rate": 3.699544764795144e-06,
+      "loss": 0.2416,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6494316032212153,
+      "grad_norm": 2.244595527648926,
+      "learning_rate": 3.690440060698028e-06,
+      "loss": 0.2002,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6502963590177815,
+      "grad_norm": 2.533815622329712,
+      "learning_rate": 3.6813353566009107e-06,
+      "loss": 0.2065,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6511611148143477,
+      "grad_norm": 1.883478045463562,
+      "learning_rate": 3.6722306525037936e-06,
+      "loss": 0.2014,
+      "step": 2259
+    },
+    {
+      "epoch": 0.652025870610914,
+      "grad_norm": 4.37358283996582,
+      "learning_rate": 3.6631259484066773e-06,
+      "loss": 0.2347,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6528906264074802,
+      "grad_norm": 1.9453434944152832,
+      "learning_rate": 3.65402124430956e-06,
+      "loss": 0.2414,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6537553822040464,
+      "grad_norm": 1.5430552959442139,
+      "learning_rate": 3.644916540212443e-06,
+      "loss": 0.2012,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6546201380006126,
+      "grad_norm": 2.023857593536377,
+      "learning_rate": 3.6358118361153266e-06,
+      "loss": 0.216,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6554848937971788,
+      "grad_norm": 2.0380475521087646,
+      "learning_rate": 3.6267071320182095e-06,
+      "loss": 0.1851,
+      "step": 2274
+    },
+    {
+      "epoch": 0.656349649593745,
+      "grad_norm": 1.9703435897827148,
+      "learning_rate": 3.617602427921093e-06,
+      "loss": 0.2152,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6572144053903112,
+      "grad_norm": 2.0153567790985107,
+      "learning_rate": 3.608497723823976e-06,
+      "loss": 0.2248,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6580791611868774,
+      "grad_norm": 1.6611104011535645,
+      "learning_rate": 3.599393019726859e-06,
+      "loss": 0.2416,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6589439169834436,
+      "grad_norm": 1.8866307735443115,
+      "learning_rate": 3.5902883156297426e-06,
+      "loss": 0.2319,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6598086727800098,
+      "grad_norm": 2.9478352069854736,
+      "learning_rate": 3.5811836115326254e-06,
+      "loss": 0.2614,
+      "step": 2289
+    },
+    {
+      "epoch": 0.660673428576576,
+      "grad_norm": 1.578539252281189,
+      "learning_rate": 3.572078907435509e-06,
+      "loss": 0.2097,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6615381843731422,
+      "grad_norm": 3.096663236618042,
+      "learning_rate": 3.562974203338392e-06,
+      "loss": 0.1977,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6624029401697084,
+      "grad_norm": 1.81285560131073,
+      "learning_rate": 3.5538694992412748e-06,
+      "loss": 0.2273,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6629794440340858,
+      "eval_loss": 0.2196592092514038,
+      "eval_mse": 0.2196592075770641,
+      "eval_runtime": 6.5661,
+      "eval_samples_per_second": 152.298,
+      "eval_steps_per_second": 19.037,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6632676959662746,
+      "grad_norm": 1.6824992895126343,
+      "learning_rate": 3.5447647951441585e-06,
+      "loss": 0.1899,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6641324517628407,
+      "grad_norm": 2.2244341373443604,
+      "learning_rate": 3.5356600910470413e-06,
+      "loss": 0.2113,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6649972075594069,
+      "grad_norm": 2.0596795082092285,
+      "learning_rate": 3.526555386949924e-06,
+      "loss": 0.2264,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6658619633559731,
+      "grad_norm": 2.2340962886810303,
+      "learning_rate": 3.517450682852808e-06,
+      "loss": 0.2168,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6667267191525393,
+      "grad_norm": 1.9558560848236084,
+      "learning_rate": 3.5083459787556907e-06,
+      "loss": 0.2088,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6675914749491055,
+      "grad_norm": 1.9379972219467163,
+      "learning_rate": 3.499241274658574e-06,
+      "loss": 0.2284,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6684562307456717,
+      "grad_norm": 2.3833818435668945,
+      "learning_rate": 3.490136570561457e-06,
+      "loss": 0.2526,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6693209865422379,
+      "grad_norm": 2.1912760734558105,
+      "learning_rate": 3.48103186646434e-06,
+      "loss": 0.2579,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6701857423388041,
+      "grad_norm": 1.4502041339874268,
+      "learning_rate": 3.4719271623672233e-06,
+      "loss": 0.2118,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6710504981353703,
+      "grad_norm": 1.6936084032058716,
+      "learning_rate": 3.4628224582701066e-06,
+      "loss": 0.2417,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6719152539319365,
+      "grad_norm": 2.5630977153778076,
+      "learning_rate": 3.45371775417299e-06,
+      "loss": 0.255,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6727800097285027,
+      "grad_norm": 1.7571030855178833,
+      "learning_rate": 3.4446130500758727e-06,
+      "loss": 0.2122,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6736447655250689,
+      "grad_norm": 2.4220032691955566,
+      "learning_rate": 3.435508345978756e-06,
+      "loss": 0.2383,
+      "step": 2337
+    },
+    {
+      "epoch": 0.6745095213216351,
+      "grad_norm": 2.3608193397521973,
+      "learning_rate": 3.4264036418816392e-06,
+      "loss": 0.241,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6753742771182013,
+      "grad_norm": 2.0964229106903076,
+      "learning_rate": 3.417298937784522e-06,
+      "loss": 0.2178,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6762390329147675,
+      "grad_norm": 3.038722038269043,
+      "learning_rate": 3.4081942336874053e-06,
+      "loss": 0.2276,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6771037887113337,
+      "grad_norm": 3.0551295280456543,
+      "learning_rate": 3.3990895295902886e-06,
+      "loss": 0.1959,
+      "step": 2349
+    },
+    {
+      "epoch": 0.6779685445078999,
+      "grad_norm": 1.4905811548233032,
+      "learning_rate": 3.3899848254931714e-06,
+      "loss": 0.1863,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6788333003044661,
+      "grad_norm": 3.0637025833129883,
+      "learning_rate": 3.380880121396055e-06,
+      "loss": 0.2341,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6796980561010323,
+      "grad_norm": 2.9851162433624268,
+      "learning_rate": 3.371775417298938e-06,
+      "loss": 0.2586,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6805628118975985,
+      "grad_norm": 2.0650391578674316,
+      "learning_rate": 3.3626707132018212e-06,
+      "loss": 0.2181,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6814275676941647,
+      "grad_norm": 3.1792430877685547,
+      "learning_rate": 3.3535660091047045e-06,
+      "loss": 0.234,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6822923234907309,
+      "grad_norm": 2.171764612197876,
+      "learning_rate": 3.3444613050075873e-06,
+      "loss": 0.2232,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6831570792872971,
+      "grad_norm": 1.8832968473434448,
+      "learning_rate": 3.3353566009104706e-06,
+      "loss": 0.2386,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6840218350838633,
+      "grad_norm": 2.217407703399658,
+      "learning_rate": 3.326251896813354e-06,
+      "loss": 0.2575,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6848865908804295,
+      "grad_norm": 1.3866760730743408,
+      "learning_rate": 3.3171471927162367e-06,
+      "loss": 0.2321,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6857513466769957,
+      "grad_norm": 2.836749315261841,
+      "learning_rate": 3.3080424886191204e-06,
+      "loss": 0.2082,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6866161024735619,
+      "grad_norm": 4.798961162567139,
+      "learning_rate": 3.2989377845220033e-06,
+      "loss": 0.2397,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6874808582701281,
+      "grad_norm": 1.8099883794784546,
+      "learning_rate": 3.289833080424886e-06,
+      "loss": 0.2468,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6883456140666943,
+      "grad_norm": 3.0821380615234375,
+      "learning_rate": 3.2807283763277698e-06,
+      "loss": 0.2013,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6892103698632605,
+      "grad_norm": 1.6952015161514282,
+      "learning_rate": 3.2716236722306526e-06,
+      "loss": 0.2156,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6900751256598266,
+      "grad_norm": 2.4413681030273438,
+      "learning_rate": 3.2625189681335363e-06,
+      "loss": 0.2307,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6909398814563928,
+      "grad_norm": 1.9589879512786865,
+      "learning_rate": 3.253414264036419e-06,
+      "loss": 0.2948,
+      "step": 2397
+    },
+    {
+      "epoch": 0.691804637252959,
+      "grad_norm": 2.4465548992156982,
+      "learning_rate": 3.244309559939302e-06,
+      "loss": 0.223,
+      "step": 2400
+    },
+    {
+      "epoch": 0.691804637252959,
+      "eval_loss": 0.22531487047672272,
+      "eval_mse": 0.2253148703626357,
+      "eval_runtime": 6.6334,
+      "eval_samples_per_second": 150.751,
+      "eval_steps_per_second": 18.844,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6926693930495252,
+      "grad_norm": 2.3037302494049072,
+      "learning_rate": 3.2352048558421857e-06,
+      "loss": 0.2199,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6935341488460914,
+      "grad_norm": 1.6721099615097046,
+      "learning_rate": 3.2261001517450685e-06,
+      "loss": 0.2286,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6943989046426576,
+      "grad_norm": 3.3806381225585938,
+      "learning_rate": 3.2169954476479514e-06,
+      "loss": 0.1988,
+      "step": 2409
+    },
+    {
+      "epoch": 0.6952636604392238,
+      "grad_norm": 2.1515021324157715,
+      "learning_rate": 3.207890743550835e-06,
+      "loss": 0.2412,
+      "step": 2412
+    },
+    {
+      "epoch": 0.69612841623579,
+      "grad_norm": 3.980482816696167,
+      "learning_rate": 3.198786039453718e-06,
+      "loss": 0.2088,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6969931720323563,
+      "grad_norm": 2.2418131828308105,
+      "learning_rate": 3.1896813353566016e-06,
+      "loss": 0.2107,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6978579278289225,
+      "grad_norm": 1.819807767868042,
+      "learning_rate": 3.1805766312594844e-06,
+      "loss": 0.2435,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6987226836254887,
+      "grad_norm": 3.8227691650390625,
+      "learning_rate": 3.1714719271623673e-06,
+      "loss": 0.2278,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6995874394220549,
+      "grad_norm": 2.207240104675293,
+      "learning_rate": 3.162367223065251e-06,
+      "loss": 0.2362,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7004521952186211,
+      "grad_norm": 1.796724796295166,
+      "learning_rate": 3.153262518968134e-06,
+      "loss": 0.2383,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7013169510151873,
+      "grad_norm": 2.7628397941589355,
+      "learning_rate": 3.1441578148710167e-06,
+      "loss": 0.2326,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7021817068117535,
+      "grad_norm": 1.3642479181289673,
+      "learning_rate": 3.1350531107739003e-06,
+      "loss": 0.2059,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7030464626083197,
+      "grad_norm": 1.5554901361465454,
+      "learning_rate": 3.125948406676783e-06,
+      "loss": 0.2162,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7039112184048859,
+      "grad_norm": 2.5311179161071777,
+      "learning_rate": 3.1168437025796665e-06,
+      "loss": 0.237,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7047759742014521,
+      "grad_norm": 1.805991768836975,
+      "learning_rate": 3.1077389984825497e-06,
+      "loss": 0.2426,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7056407299980183,
+      "grad_norm": 2.029891014099121,
+      "learning_rate": 3.0986342943854326e-06,
+      "loss": 0.2139,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7065054857945845,
+      "grad_norm": 2.5495572090148926,
+      "learning_rate": 3.089529590288316e-06,
+      "loss": 0.2279,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7073702415911507,
+      "grad_norm": 1.5610768795013428,
+      "learning_rate": 3.080424886191199e-06,
+      "loss": 0.2175,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7082349973877169,
+      "grad_norm": 3.5896153450012207,
+      "learning_rate": 3.0713201820940824e-06,
+      "loss": 0.2459,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7090997531842831,
+      "grad_norm": 2.1120688915252686,
+      "learning_rate": 3.062215477996965e-06,
+      "loss": 0.2565,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7099645089808493,
+      "grad_norm": 2.7099833488464355,
+      "learning_rate": 3.0531107738998485e-06,
+      "loss": 0.2088,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7108292647774155,
+      "grad_norm": 1.7153905630111694,
+      "learning_rate": 3.0440060698027317e-06,
+      "loss": 0.2114,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7116940205739817,
+      "grad_norm": 1.9465105533599854,
+      "learning_rate": 3.0349013657056146e-06,
+      "loss": 0.1932,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7125587763705479,
+      "grad_norm": 1.6483453512191772,
+      "learning_rate": 3.025796661608498e-06,
+      "loss": 0.232,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7134235321671141,
+      "grad_norm": 2.3854711055755615,
+      "learning_rate": 3.016691957511381e-06,
+      "loss": 0.2397,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7142882879636803,
+      "grad_norm": 1.565765380859375,
+      "learning_rate": 3.0075872534142644e-06,
+      "loss": 0.2055,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7151530437602465,
+      "grad_norm": 1.5985909700393677,
+      "learning_rate": 2.9984825493171476e-06,
+      "loss": 0.2077,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7160177995568127,
+      "grad_norm": 2.9907102584838867,
+      "learning_rate": 2.9893778452200305e-06,
+      "loss": 0.2125,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7168825553533789,
+      "grad_norm": 3.0764994621276855,
+      "learning_rate": 2.9802731411229137e-06,
+      "loss": 0.2342,
+      "step": 2487
+    },
+    {
+      "epoch": 0.717747311149945,
+      "grad_norm": 2.954237461090088,
+      "learning_rate": 2.971168437025797e-06,
+      "loss": 0.232,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7186120669465113,
+      "grad_norm": 1.5421547889709473,
+      "learning_rate": 2.96206373292868e-06,
+      "loss": 0.2019,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7194768227430774,
+      "grad_norm": 5.054042816162109,
+      "learning_rate": 2.9529590288315635e-06,
+      "loss": 0.2359,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7203415785396436,
+      "grad_norm": 1.5477067232131958,
+      "learning_rate": 2.9438543247344464e-06,
+      "loss": 0.2028,
+      "step": 2499
+    },
+    {
+      "epoch": 0.7206298304718324,
+      "eval_loss": 0.22387926280498505,
+      "eval_mse": 0.22387926151184365,
+      "eval_runtime": 6.5037,
+      "eval_samples_per_second": 153.759,
+      "eval_steps_per_second": 19.22,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7212063343362098,
+      "grad_norm": 1.912636160850525,
+      "learning_rate": 2.9347496206373292e-06,
+      "loss": 0.2167,
+      "step": 2502
+    },
+    {
+      "epoch": 0.722071090132776,
+      "grad_norm": 1.644394040107727,
+      "learning_rate": 2.925644916540213e-06,
+      "loss": 0.2068,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7229358459293422,
+      "grad_norm": 3.4328315258026123,
+      "learning_rate": 2.9165402124430958e-06,
+      "loss": 0.2417,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7238006017259084,
+      "grad_norm": 1.918043613433838,
+      "learning_rate": 2.9074355083459786e-06,
+      "loss": 0.2308,
+      "step": 2511
+    },
+    {
+      "epoch": 0.7246653575224746,
+      "grad_norm": 1.94221031665802,
+      "learning_rate": 2.8983308042488623e-06,
+      "loss": 0.2957,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7255301133190408,
+      "grad_norm": 2.877037525177002,
+      "learning_rate": 2.889226100151745e-06,
+      "loss": 0.2133,
+      "step": 2517
+    },
+    {
+      "epoch": 0.726394869115607,
+      "grad_norm": 2.2768120765686035,
+      "learning_rate": 2.880121396054629e-06,
+      "loss": 0.2251,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7272596249121732,
+      "grad_norm": 2.9239742755889893,
+      "learning_rate": 2.8710166919575117e-06,
+      "loss": 0.215,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7281243807087394,
+      "grad_norm": 1.5520339012145996,
+      "learning_rate": 2.8619119878603945e-06,
+      "loss": 0.2113,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7289891365053056,
+      "grad_norm": 3.458822011947632,
+      "learning_rate": 2.852807283763278e-06,
+      "loss": 0.2465,
+      "step": 2529
+    },
+    {
+      "epoch": 0.7298538923018718,
+      "grad_norm": 1.606724500656128,
+      "learning_rate": 2.843702579666161e-06,
+      "loss": 0.188,
+      "step": 2532
+    },
+    {
+      "epoch": 0.730718648098438,
+      "grad_norm": 3.552236318588257,
+      "learning_rate": 2.834597875569044e-06,
+      "loss": 0.2178,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7315834038950042,
+      "grad_norm": 2.940363883972168,
+      "learning_rate": 2.8254931714719276e-06,
+      "loss": 0.2149,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7324481596915704,
+      "grad_norm": 1.3787034749984741,
+      "learning_rate": 2.8163884673748104e-06,
+      "loss": 0.2155,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7333129154881366,
+      "grad_norm": 1.4637516736984253,
+      "learning_rate": 2.807283763277694e-06,
+      "loss": 0.1944,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7341776712847028,
+      "grad_norm": 1.5903189182281494,
+      "learning_rate": 2.798179059180577e-06,
+      "loss": 0.214,
+      "step": 2547
+    },
+    {
+      "epoch": 0.735042427081269,
+      "grad_norm": 1.999664306640625,
+      "learning_rate": 2.78907435508346e-06,
+      "loss": 0.2277,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7359071828778352,
+      "grad_norm": 2.250450849533081,
+      "learning_rate": 2.7799696509863435e-06,
+      "loss": 0.1958,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7367719386744014,
+      "grad_norm": 1.8212629556655884,
+      "learning_rate": 2.7708649468892263e-06,
+      "loss": 0.2155,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7376366944709676,
+      "grad_norm": 1.4670761823654175,
+      "learning_rate": 2.7617602427921096e-06,
+      "loss": 0.2251,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7385014502675338,
+      "grad_norm": 2.9052860736846924,
+      "learning_rate": 2.752655538694993e-06,
+      "loss": 0.2279,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7393662060641,
+      "grad_norm": 1.647455096244812,
+      "learning_rate": 2.7435508345978757e-06,
+      "loss": 0.2011,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7402309618606662,
+      "grad_norm": 2.5667457580566406,
+      "learning_rate": 2.734446130500759e-06,
+      "loss": 0.2113,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7410957176572324,
+      "grad_norm": 2.021571159362793,
+      "learning_rate": 2.7253414264036422e-06,
+      "loss": 0.2085,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7419604734537987,
+      "grad_norm": 3.436924457550049,
+      "learning_rate": 2.716236722306525e-06,
+      "loss": 0.2658,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7428252292503649,
+      "grad_norm": 1.9480434656143188,
+      "learning_rate": 2.7071320182094083e-06,
+      "loss": 0.2595,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7436899850469311,
+      "grad_norm": 1.8556321859359741,
+      "learning_rate": 2.6980273141122916e-06,
+      "loss": 0.2308,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7445547408434973,
+      "grad_norm": 3.174111843109131,
+      "learning_rate": 2.688922610015175e-06,
+      "loss": 0.2172,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7454194966400635,
+      "grad_norm": 1.5629518032073975,
+      "learning_rate": 2.6798179059180577e-06,
+      "loss": 0.2417,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7462842524366297,
+      "grad_norm": 1.8133536577224731,
+      "learning_rate": 2.670713201820941e-06,
+      "loss": 0.2188,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7471490082331959,
+      "grad_norm": 1.5448634624481201,
+      "learning_rate": 2.6616084977238242e-06,
+      "loss": 0.2201,
+      "step": 2592
+    },
+    {
+      "epoch": 0.748013764029762,
+      "grad_norm": 2.376194953918457,
+      "learning_rate": 2.652503793626707e-06,
+      "loss": 0.2146,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7488785198263282,
+      "grad_norm": 1.8988285064697266,
+      "learning_rate": 2.6433990895295904e-06,
+      "loss": 0.2322,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7494550236907057,
+      "eval_loss": 0.21797478199005127,
+      "eval_mse": 0.21797478066571058,
+      "eval_runtime": 6.533,
+      "eval_samples_per_second": 153.07,
+      "eval_steps_per_second": 19.134,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7497432756228944,
+      "grad_norm": 3.16768217086792,
+      "learning_rate": 2.6342943854324736e-06,
+      "loss": 0.2533,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7506080314194606,
+      "grad_norm": 1.7228988409042358,
+      "learning_rate": 2.625189681335357e-06,
+      "loss": 0.2406,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7514727872160268,
+      "grad_norm": 2.9629013538360596,
+      "learning_rate": 2.61608497723824e-06,
+      "loss": 0.2131,
+      "step": 2607
+    },
+    {
+      "epoch": 0.752337543012593,
+      "grad_norm": 3.4181559085845947,
+      "learning_rate": 2.606980273141123e-06,
+      "loss": 0.2126,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7532022988091592,
+      "grad_norm": 2.142685890197754,
+      "learning_rate": 2.5978755690440063e-06,
+      "loss": 0.2345,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7540670546057254,
+      "grad_norm": 3.622145175933838,
+      "learning_rate": 2.5887708649468895e-06,
+      "loss": 0.2187,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7549318104022916,
+      "grad_norm": 1.5243175029754639,
+      "learning_rate": 2.5796661608497724e-06,
+      "loss": 0.1996,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7557965661988578,
+      "grad_norm": 2.7075355052948,
+      "learning_rate": 2.570561456752656e-06,
+      "loss": 0.2755,
+      "step": 2622
+    },
+    {
+      "epoch": 0.756661321995424,
+      "grad_norm": 2.6778082847595215,
+      "learning_rate": 2.561456752655539e-06,
+      "loss": 0.1959,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7575260777919902,
+      "grad_norm": 3.8043668270111084,
+      "learning_rate": 2.5523520485584217e-06,
+      "loss": 0.2054,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7583908335885564,
+      "grad_norm": 3.628281354904175,
+      "learning_rate": 2.5432473444613054e-06,
+      "loss": 0.2095,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7592555893851226,
+      "grad_norm": 3.2685089111328125,
+      "learning_rate": 2.5341426403641883e-06,
+      "loss": 0.2167,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7601203451816888,
+      "grad_norm": 1.6048041582107544,
+      "learning_rate": 2.525037936267071e-06,
+      "loss": 0.2416,
+      "step": 2637
+    },
+    {
+      "epoch": 0.760985100978255,
+      "grad_norm": 1.5175867080688477,
+      "learning_rate": 2.515933232169955e-06,
+      "loss": 0.2326,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7618498567748212,
+      "grad_norm": 1.9787158966064453,
+      "learning_rate": 2.5068285280728376e-06,
+      "loss": 0.2266,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7627146125713874,
+      "grad_norm": 4.297834873199463,
+      "learning_rate": 2.497723823975721e-06,
+      "loss": 0.2414,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7635793683679536,
+      "grad_norm": 1.822587251663208,
+      "learning_rate": 2.488619119878604e-06,
+      "loss": 0.2386,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7644441241645198,
+      "grad_norm": 4.966648101806641,
+      "learning_rate": 2.4795144157814874e-06,
+      "loss": 0.2268,
+      "step": 2652
+    },
+    {
+      "epoch": 0.765308879961086,
+      "grad_norm": 2.8137335777282715,
+      "learning_rate": 2.4704097116843703e-06,
+      "loss": 0.2349,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7661736357576522,
+      "grad_norm": 2.5401577949523926,
+      "learning_rate": 2.4613050075872536e-06,
+      "loss": 0.1941,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7670383915542184,
+      "grad_norm": 1.550758957862854,
+      "learning_rate": 2.452200303490137e-06,
+      "loss": 0.2151,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7679031473507846,
+      "grad_norm": 6.889467239379883,
+      "learning_rate": 2.44309559939302e-06,
+      "loss": 0.2287,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7687679031473508,
+      "grad_norm": 1.4833314418792725,
+      "learning_rate": 2.4339908952959034e-06,
+      "loss": 0.2017,
+      "step": 2667
+    },
+    {
+      "epoch": 0.769632658943917,
+      "grad_norm": 1.607751727104187,
+      "learning_rate": 2.424886191198786e-06,
+      "loss": 0.2309,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7704974147404832,
+      "grad_norm": 1.7369478940963745,
+      "learning_rate": 2.4157814871016695e-06,
+      "loss": 0.217,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7713621705370494,
+      "grad_norm": 1.7309290170669556,
+      "learning_rate": 2.4066767830045527e-06,
+      "loss": 0.2151,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7722269263336156,
+      "grad_norm": 2.047727108001709,
+      "learning_rate": 2.397572078907436e-06,
+      "loss": 0.2078,
+      "step": 2679
+    },
+    {
+      "epoch": 0.7730916821301818,
+      "grad_norm": 2.2800142765045166,
+      "learning_rate": 2.388467374810319e-06,
+      "loss": 0.2372,
+      "step": 2682
+    },
+    {
+      "epoch": 0.773956437926748,
+      "grad_norm": 3.920849323272705,
+      "learning_rate": 2.379362670713202e-06,
+      "loss": 0.2304,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7748211937233141,
+      "grad_norm": 3.4216678142547607,
+      "learning_rate": 2.3702579666160854e-06,
+      "loss": 0.2382,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7756859495198803,
+      "grad_norm": 1.5471861362457275,
+      "learning_rate": 2.3611532625189686e-06,
+      "loss": 0.2243,
+      "step": 2691
+    },
+    {
+      "epoch": 0.7765507053164465,
+      "grad_norm": 1.7489866018295288,
+      "learning_rate": 2.3520485584218515e-06,
+      "loss": 0.2239,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7774154611130127,
+      "grad_norm": 4.3836822509765625,
+      "learning_rate": 2.3429438543247347e-06,
+      "loss": 0.205,
+      "step": 2697
+    },
+    {
+      "epoch": 0.7782802169095789,
+      "grad_norm": 1.8707342147827148,
+      "learning_rate": 2.333839150227618e-06,
+      "loss": 0.1933,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7782802169095789,
+      "eval_loss": 0.2158193737268448,
+      "eval_mse": 0.21581936262454837,
+      "eval_runtime": 6.5614,
+      "eval_samples_per_second": 152.407,
+      "eval_steps_per_second": 19.051,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7791449727061451,
+      "grad_norm": 1.6554896831512451,
+      "learning_rate": 2.324734446130501e-06,
+      "loss": 0.2178,
+      "step": 2703
+    },
+    {
+      "epoch": 0.7800097285027113,
+      "grad_norm": 3.131352424621582,
+      "learning_rate": 2.315629742033384e-06,
+      "loss": 0.2122,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7808744842992775,
+      "grad_norm": 2.62422776222229,
+      "learning_rate": 2.3065250379362674e-06,
+      "loss": 0.2085,
+      "step": 2709
+    },
+    {
+      "epoch": 0.7817392400958437,
+      "grad_norm": 2.1258456707000732,
+      "learning_rate": 2.2974203338391502e-06,
+      "loss": 0.2203,
+      "step": 2712
+    },
+    {
+      "epoch": 0.7826039958924099,
+      "grad_norm": 3.144688606262207,
+      "learning_rate": 2.2883156297420335e-06,
+      "loss": 0.2428,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7834687516889761,
+      "grad_norm": 4.5740180015563965,
+      "learning_rate": 2.2792109256449168e-06,
+      "loss": 0.2196,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7843335074855423,
+      "grad_norm": 1.7256083488464355,
+      "learning_rate": 2.2701062215477996e-06,
+      "loss": 0.2316,
+      "step": 2721
+    },
+    {
+      "epoch": 0.7851982632821085,
+      "grad_norm": 2.0723230838775635,
+      "learning_rate": 2.261001517450683e-06,
+      "loss": 0.1958,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7860630190786747,
+      "grad_norm": 1.6268962621688843,
+      "learning_rate": 2.251896813353566e-06,
+      "loss": 0.2329,
+      "step": 2727
+    },
+    {
+      "epoch": 0.786927774875241,
+      "grad_norm": 4.054417610168457,
+      "learning_rate": 2.2427921092564494e-06,
+      "loss": 0.2091,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7877925306718072,
+      "grad_norm": 1.7409260272979736,
+      "learning_rate": 2.2336874051593322e-06,
+      "loss": 0.2144,
+      "step": 2733
+    },
+    {
+      "epoch": 0.7886572864683734,
+      "grad_norm": 2.77607798576355,
+      "learning_rate": 2.2245827010622155e-06,
+      "loss": 0.2199,
+      "step": 2736
+    },
+    {
+      "epoch": 0.7895220422649396,
+      "grad_norm": 2.215284585952759,
+      "learning_rate": 2.2154779969650988e-06,
+      "loss": 0.2388,
+      "step": 2739
+    },
+    {
+      "epoch": 0.7903867980615058,
+      "grad_norm": 1.7318382263183594,
+      "learning_rate": 2.206373292867982e-06,
+      "loss": 0.2115,
+      "step": 2742
+    },
+    {
+      "epoch": 0.791251553858072,
+      "grad_norm": 1.5627691745758057,
+      "learning_rate": 2.197268588770865e-06,
+      "loss": 0.2054,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7921163096546382,
+      "grad_norm": 1.810509443283081,
+      "learning_rate": 2.188163884673748e-06,
+      "loss": 0.2086,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7929810654512044,
+      "grad_norm": 2.1531972885131836,
+      "learning_rate": 2.1790591805766314e-06,
+      "loss": 0.2252,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7938458212477706,
+      "grad_norm": 2.0212440490722656,
+      "learning_rate": 2.1699544764795147e-06,
+      "loss": 0.2355,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7947105770443368,
+      "grad_norm": 5.030855178833008,
+      "learning_rate": 2.1608497723823975e-06,
+      "loss": 0.2341,
+      "step": 2757
+    },
+    {
+      "epoch": 0.795575332840903,
+      "grad_norm": 2.213249921798706,
+      "learning_rate": 2.1517450682852808e-06,
+      "loss": 0.2133,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7964400886374692,
+      "grad_norm": 1.8025689125061035,
+      "learning_rate": 2.142640364188164e-06,
+      "loss": 0.2251,
+      "step": 2763
+    },
+    {
+      "epoch": 0.7973048444340354,
+      "grad_norm": 4.14149284362793,
+      "learning_rate": 2.1335356600910473e-06,
+      "loss": 0.253,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7981696002306016,
+      "grad_norm": 2.2051069736480713,
+      "learning_rate": 2.12443095599393e-06,
+      "loss": 0.2238,
+      "step": 2769
+    },
+    {
+      "epoch": 0.7990343560271678,
+      "grad_norm": 2.249032497406006,
+      "learning_rate": 2.1153262518968134e-06,
+      "loss": 0.2282,
+      "step": 2772
+    },
+    {
+      "epoch": 0.799899111823734,
+      "grad_norm": 1.5087867975234985,
+      "learning_rate": 2.1062215477996967e-06,
+      "loss": 0.1948,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8007638676203002,
+      "grad_norm": 1.9934585094451904,
+      "learning_rate": 2.09711684370258e-06,
+      "loss": 0.2061,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8016286234168664,
+      "grad_norm": 2.521526336669922,
+      "learning_rate": 2.0880121396054632e-06,
+      "loss": 0.2331,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8024933792134326,
+      "grad_norm": 4.441010475158691,
+      "learning_rate": 2.078907435508346e-06,
+      "loss": 0.2337,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8033581350099988,
+      "grad_norm": 1.9386543035507202,
+      "learning_rate": 2.0698027314112293e-06,
+      "loss": 0.2434,
+      "step": 2787
+    },
+    {
+      "epoch": 0.804222890806565,
+      "grad_norm": 1.6140722036361694,
+      "learning_rate": 2.0606980273141126e-06,
+      "loss": 0.223,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8050876466031311,
+      "grad_norm": 3.248769998550415,
+      "learning_rate": 2.051593323216996e-06,
+      "loss": 0.2352,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8059524023996973,
+      "grad_norm": 2.259561061859131,
+      "learning_rate": 2.0424886191198787e-06,
+      "loss": 0.2428,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8068171581962635,
+      "grad_norm": 2.289113998413086,
+      "learning_rate": 2.033383915022762e-06,
+      "loss": 0.2085,
+      "step": 2799
+    },
+    {
+      "epoch": 0.8071054101284523,
+      "eval_loss": 0.22976131737232208,
+      "eval_mse": 0.22976131996285404,
+      "eval_runtime": 6.5104,
+      "eval_samples_per_second": 153.6,
+      "eval_steps_per_second": 19.2,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8076819139928297,
+      "grad_norm": 1.4588847160339355,
+      "learning_rate": 2.0242792109256452e-06,
+      "loss": 0.2004,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8085466697893959,
+      "grad_norm": 3.1680517196655273,
+      "learning_rate": 2.0151745068285285e-06,
+      "loss": 0.2038,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8094114255859621,
+      "grad_norm": 2.708411693572998,
+      "learning_rate": 2.0060698027314113e-06,
+      "loss": 0.2289,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8102761813825283,
+      "grad_norm": 2.2479403018951416,
+      "learning_rate": 1.9969650986342946e-06,
+      "loss": 0.2465,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8111409371790945,
+      "grad_norm": 3.2582664489746094,
+      "learning_rate": 1.987860394537178e-06,
+      "loss": 0.2187,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8120056929756607,
+      "grad_norm": 2.5267367362976074,
+      "learning_rate": 1.978755690440061e-06,
+      "loss": 0.1955,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8128704487722269,
+      "grad_norm": 2.42645525932312,
+      "learning_rate": 1.969650986342944e-06,
+      "loss": 0.233,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8137352045687931,
+      "grad_norm": 1.6183414459228516,
+      "learning_rate": 1.9605462822458273e-06,
+      "loss": 0.2187,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8145999603653593,
+      "grad_norm": 2.7215240001678467,
+      "learning_rate": 1.9514415781487105e-06,
+      "loss": 0.2109,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8154647161619255,
+      "grad_norm": 2.152639389038086,
+      "learning_rate": 1.9423368740515934e-06,
+      "loss": 0.2056,
+      "step": 2829
+    },
+    {
+      "epoch": 0.8163294719584917,
+      "grad_norm": 2.530045509338379,
+      "learning_rate": 1.9332321699544766e-06,
+      "loss": 0.2185,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8171942277550579,
+      "grad_norm": 1.8189818859100342,
+      "learning_rate": 1.92412746585736e-06,
+      "loss": 0.2502,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8180589835516241,
+      "grad_norm": 1.5473167896270752,
+      "learning_rate": 1.9150227617602427e-06,
+      "loss": 0.1923,
+      "step": 2838
+    },
+    {
+      "epoch": 0.8189237393481903,
+      "grad_norm": 2.99245285987854,
+      "learning_rate": 1.9059180576631262e-06,
+      "loss": 0.209,
+      "step": 2841
+    },
+    {
+      "epoch": 0.8197884951447565,
+      "grad_norm": 1.509892225265503,
+      "learning_rate": 1.8968133535660093e-06,
+      "loss": 0.2293,
+      "step": 2844
+    },
+    {
+      "epoch": 0.8206532509413227,
+      "grad_norm": 3.5667645931243896,
+      "learning_rate": 1.8877086494688923e-06,
+      "loss": 0.1935,
+      "step": 2847
+    },
+    {
+      "epoch": 0.8215180067378889,
+      "grad_norm": 3.1929867267608643,
+      "learning_rate": 1.8786039453717756e-06,
+      "loss": 0.201,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8223827625344551,
+      "grad_norm": 1.5706427097320557,
+      "learning_rate": 1.8694992412746589e-06,
+      "loss": 0.2315,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8232475183310213,
+      "grad_norm": 3.1730329990386963,
+      "learning_rate": 1.860394537177542e-06,
+      "loss": 0.2475,
+      "step": 2856
+    },
+    {
+      "epoch": 0.8241122741275875,
+      "grad_norm": 1.3369548320770264,
+      "learning_rate": 1.851289833080425e-06,
+      "loss": 0.2186,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8249770299241537,
+      "grad_norm": 2.522751569747925,
+      "learning_rate": 1.8421851289833082e-06,
+      "loss": 0.2477,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8258417857207199,
+      "grad_norm": 1.991076111793518,
+      "learning_rate": 1.8330804248861913e-06,
+      "loss": 0.2009,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8267065415172861,
+      "grad_norm": 2.707282781600952,
+      "learning_rate": 1.8239757207890745e-06,
+      "loss": 0.2302,
+      "step": 2868
+    },
+    {
+      "epoch": 0.8275712973138523,
+      "grad_norm": 1.578192949295044,
+      "learning_rate": 1.8148710166919576e-06,
+      "loss": 0.2021,
+      "step": 2871
+    },
+    {
+      "epoch": 0.8284360531104185,
+      "grad_norm": 3.6003148555755615,
+      "learning_rate": 1.8057663125948407e-06,
+      "loss": 0.2409,
+      "step": 2874
+    },
+    {
+      "epoch": 0.8293008089069847,
+      "grad_norm": 3.1442506313323975,
+      "learning_rate": 1.796661608497724e-06,
+      "loss": 0.2129,
+      "step": 2877
+    },
+    {
+      "epoch": 0.8301655647035509,
+      "grad_norm": 1.509333610534668,
+      "learning_rate": 1.7875569044006072e-06,
+      "loss": 0.2202,
+      "step": 2880
+    },
+    {
+      "epoch": 0.831030320500117,
+      "grad_norm": 1.9379024505615234,
+      "learning_rate": 1.7784522003034905e-06,
+      "loss": 0.2238,
+      "step": 2883
+    },
+    {
+      "epoch": 0.8318950762966834,
+      "grad_norm": 1.3617918491363525,
+      "learning_rate": 1.7693474962063733e-06,
+      "loss": 0.1972,
+      "step": 2886
+    },
+    {
+      "epoch": 0.8327598320932496,
+      "grad_norm": 1.4775515794754028,
+      "learning_rate": 1.7602427921092566e-06,
+      "loss": 0.1985,
+      "step": 2889
+    },
+    {
+      "epoch": 0.8336245878898157,
+      "grad_norm": 1.4302202463150024,
+      "learning_rate": 1.7511380880121398e-06,
+      "loss": 0.2403,
+      "step": 2892
+    },
+    {
+      "epoch": 0.834489343686382,
+      "grad_norm": 2.128401041030884,
+      "learning_rate": 1.742033383915023e-06,
+      "loss": 0.2204,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8353540994829481,
+      "grad_norm": 1.9766216278076172,
+      "learning_rate": 1.732928679817906e-06,
+      "loss": 0.2038,
+      "step": 2898
+    },
+    {
+      "epoch": 0.8359306033473256,
+      "eval_loss": 0.21660968661308289,
+      "eval_mse": 0.21660968138270925,
+      "eval_runtime": 6.5731,
+      "eval_samples_per_second": 152.135,
+      "eval_steps_per_second": 19.017,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8362188552795143,
+      "grad_norm": 2.260625123977661,
+      "learning_rate": 1.7238239757207892e-06,
+      "loss": 0.2206,
+      "step": 2901
+    },
+    {
+      "epoch": 0.8370836110760805,
+      "grad_norm": 2.605224609375,
+      "learning_rate": 1.7147192716236725e-06,
+      "loss": 0.2186,
+      "step": 2904
+    },
+    {
+      "epoch": 0.8379483668726467,
+      "grad_norm": 2.7619729042053223,
+      "learning_rate": 1.7056145675265557e-06,
+      "loss": 0.2195,
+      "step": 2907
+    },
+    {
+      "epoch": 0.8388131226692129,
+      "grad_norm": 1.9395873546600342,
+      "learning_rate": 1.6965098634294386e-06,
+      "loss": 0.2119,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8396778784657791,
+      "grad_norm": 1.727263331413269,
+      "learning_rate": 1.6874051593323218e-06,
+      "loss": 0.2125,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8405426342623453,
+      "grad_norm": 2.149775981903076,
+      "learning_rate": 1.6783004552352051e-06,
+      "loss": 0.2644,
+      "step": 2916
+    },
+    {
+      "epoch": 0.8414073900589115,
+      "grad_norm": 2.6743104457855225,
+      "learning_rate": 1.6691957511380882e-06,
+      "loss": 0.2217,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8422721458554777,
+      "grad_norm": 2.795736074447632,
+      "learning_rate": 1.6600910470409712e-06,
+      "loss": 0.2115,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8431369016520439,
+      "grad_norm": 1.8719727993011475,
+      "learning_rate": 1.6509863429438545e-06,
+      "loss": 0.1768,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8440016574486101,
+      "grad_norm": 3.4366025924682617,
+      "learning_rate": 1.6418816388467375e-06,
+      "loss": 0.2357,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8448664132451763,
+      "grad_norm": 2.2458267211914062,
+      "learning_rate": 1.6327769347496208e-06,
+      "loss": 0.2147,
+      "step": 2931
+    },
+    {
+      "epoch": 0.8457311690417425,
+      "grad_norm": 1.958115577697754,
+      "learning_rate": 1.6236722306525039e-06,
+      "loss": 0.2339,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8465959248383087,
+      "grad_norm": 1.6470586061477661,
+      "learning_rate": 1.614567526555387e-06,
+      "loss": 0.2259,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8474606806348749,
+      "grad_norm": 1.2936792373657227,
+      "learning_rate": 1.6054628224582702e-06,
+      "loss": 0.1981,
+      "step": 2940
+    },
+    {
+      "epoch": 0.8483254364314411,
+      "grad_norm": 4.565530300140381,
+      "learning_rate": 1.5963581183611534e-06,
+      "loss": 0.1993,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8491901922280073,
+      "grad_norm": 2.8682401180267334,
+      "learning_rate": 1.5872534142640367e-06,
+      "loss": 0.2093,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8500549480245735,
+      "grad_norm": 1.7801469564437866,
+      "learning_rate": 1.5781487101669196e-06,
+      "loss": 0.2136,
+      "step": 2949
+    },
+    {
+      "epoch": 0.8509197038211397,
+      "grad_norm": 2.372549057006836,
+      "learning_rate": 1.5690440060698028e-06,
+      "loss": 0.1845,
+      "step": 2952
+    },
+    {
+      "epoch": 0.8517844596177059,
+      "grad_norm": 2.190469741821289,
+      "learning_rate": 1.559939301972686e-06,
+      "loss": 0.2137,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8526492154142721,
+      "grad_norm": 1.6399952173233032,
+      "learning_rate": 1.5508345978755694e-06,
+      "loss": 0.2206,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8535139712108383,
+      "grad_norm": 1.6555943489074707,
+      "learning_rate": 1.5417298937784522e-06,
+      "loss": 0.2403,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8543787270074045,
+      "grad_norm": 2.6609280109405518,
+      "learning_rate": 1.5326251896813355e-06,
+      "loss": 0.2176,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8552434828039707,
+      "grad_norm": 2.3398261070251465,
+      "learning_rate": 1.5235204855842187e-06,
+      "loss": 0.219,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8561082386005369,
+      "grad_norm": 1.9740712642669678,
+      "learning_rate": 1.514415781487102e-06,
+      "loss": 0.2232,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8569729943971031,
+      "grad_norm": 1.6300252676010132,
+      "learning_rate": 1.5053110773899848e-06,
+      "loss": 0.2094,
+      "step": 2973
+    },
+    {
+      "epoch": 0.8578377501936693,
+      "grad_norm": 2.8211612701416016,
+      "learning_rate": 1.496206373292868e-06,
+      "loss": 0.1696,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8587025059902355,
+      "grad_norm": 2.621321439743042,
+      "learning_rate": 1.4871016691957514e-06,
+      "loss": 0.2357,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8595672617868017,
+      "grad_norm": 1.5020322799682617,
+      "learning_rate": 1.4779969650986344e-06,
+      "loss": 0.2456,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8604320175833678,
+      "grad_norm": 1.474507212638855,
+      "learning_rate": 1.4688922610015175e-06,
+      "loss": 0.2074,
+      "step": 2985
+    },
+    {
+      "epoch": 0.861296773379934,
+      "grad_norm": 2.9856317043304443,
+      "learning_rate": 1.4597875569044007e-06,
+      "loss": 0.2241,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8621615291765002,
+      "grad_norm": 2.0011954307556152,
+      "learning_rate": 1.4506828528072838e-06,
+      "loss": 0.2026,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8630262849730664,
+      "grad_norm": 1.6045671701431274,
+      "learning_rate": 1.441578148710167e-06,
+      "loss": 0.186,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8638910407696326,
+      "grad_norm": 1.5708575248718262,
+      "learning_rate": 1.4324734446130503e-06,
+      "loss": 0.2048,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8647557965661988,
+      "grad_norm": 2.8704543113708496,
+      "learning_rate": 1.4233687405159332e-06,
+      "loss": 0.2158,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8647557965661988,
+      "eval_loss": 0.2083793729543686,
+      "eval_mse": 0.20837937731285638,
+      "eval_runtime": 6.6796,
+      "eval_samples_per_second": 149.709,
+      "eval_steps_per_second": 18.714,
+      "step": 3000
+    },
+    {
+      "epoch": 0.865620552362765,
+      "grad_norm": 3.2445404529571533,
+      "learning_rate": 1.4142640364188164e-06,
+      "loss": 0.2144,
+      "step": 3003
+    },
+    {
+      "epoch": 0.8664853081593312,
+      "grad_norm": 1.96418297290802,
+      "learning_rate": 1.4051593323216997e-06,
+      "loss": 0.2491,
+      "step": 3006
+    },
+    {
+      "epoch": 0.8673500639558974,
+      "grad_norm": 1.8195468187332153,
+      "learning_rate": 1.396054628224583e-06,
+      "loss": 0.2065,
+      "step": 3009
+    },
+    {
+      "epoch": 0.8682148197524636,
+      "grad_norm": 1.3888121843338013,
+      "learning_rate": 1.3869499241274658e-06,
+      "loss": 0.2049,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8690795755490298,
+      "grad_norm": 3.738133668899536,
+      "learning_rate": 1.377845220030349e-06,
+      "loss": 0.2522,
+      "step": 3015
+    },
+    {
+      "epoch": 0.869944331345596,
+      "grad_norm": 6.097411632537842,
+      "learning_rate": 1.3687405159332323e-06,
+      "loss": 0.2532,
+      "step": 3018
+    },
+    {
+      "epoch": 0.8708090871421622,
+      "grad_norm": 2.2873427867889404,
+      "learning_rate": 1.3596358118361156e-06,
+      "loss": 0.2382,
+      "step": 3021
+    },
+    {
+      "epoch": 0.8716738429387284,
+      "grad_norm": 1.540143370628357,
+      "learning_rate": 1.3505311077389985e-06,
+      "loss": 0.1979,
+      "step": 3024
+    },
+    {
+      "epoch": 0.8725385987352946,
+      "grad_norm": 1.6231845617294312,
+      "learning_rate": 1.3414264036418817e-06,
+      "loss": 0.195,
+      "step": 3027
+    },
+    {
+      "epoch": 0.8734033545318608,
+      "grad_norm": 2.2970290184020996,
+      "learning_rate": 1.332321699544765e-06,
+      "loss": 0.2154,
+      "step": 3030
+    },
+    {
+      "epoch": 0.874268110328427,
+      "grad_norm": 2.0112340450286865,
+      "learning_rate": 1.3232169954476482e-06,
+      "loss": 0.2355,
+      "step": 3033
+    },
+    {
+      "epoch": 0.8751328661249932,
+      "grad_norm": 1.3887783288955688,
+      "learning_rate": 1.314112291350531e-06,
+      "loss": 0.201,
+      "step": 3036
+    },
+    {
+      "epoch": 0.8759976219215594,
+      "grad_norm": 2.187082529067993,
+      "learning_rate": 1.3050075872534144e-06,
+      "loss": 0.1965,
+      "step": 3039
+    },
+    {
+      "epoch": 0.8768623777181257,
+      "grad_norm": 1.300243616104126,
+      "learning_rate": 1.2959028831562976e-06,
+      "loss": 0.2097,
+      "step": 3042
+    },
+    {
+      "epoch": 0.8777271335146919,
+      "grad_norm": 2.1217234134674072,
+      "learning_rate": 1.2867981790591807e-06,
+      "loss": 0.214,
+      "step": 3045
+    },
+    {
+      "epoch": 0.8785918893112581,
+      "grad_norm": 1.8281973600387573,
+      "learning_rate": 1.277693474962064e-06,
+      "loss": 0.2082,
+      "step": 3048
+    },
+    {
+      "epoch": 0.8794566451078243,
+      "grad_norm": 2.3602306842803955,
+      "learning_rate": 1.268588770864947e-06,
+      "loss": 0.2172,
+      "step": 3051
+    },
+    {
+      "epoch": 0.8803214009043905,
+      "grad_norm": 1.903954267501831,
+      "learning_rate": 1.25948406676783e-06,
+      "loss": 0.2323,
+      "step": 3054
+    },
+    {
+      "epoch": 0.8811861567009567,
+      "grad_norm": 3.514057159423828,
+      "learning_rate": 1.2503793626707133e-06,
+      "loss": 0.2323,
+      "step": 3057
+    },
+    {
+      "epoch": 0.8820509124975229,
+      "grad_norm": 3.3089487552642822,
+      "learning_rate": 1.2412746585735964e-06,
+      "loss": 0.2506,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8829156682940891,
+      "grad_norm": 2.317981004714966,
+      "learning_rate": 1.2321699544764796e-06,
+      "loss": 0.2274,
+      "step": 3063
+    },
+    {
+      "epoch": 0.8837804240906553,
+      "grad_norm": 2.7326478958129883,
+      "learning_rate": 1.2230652503793627e-06,
+      "loss": 0.2057,
+      "step": 3066
+    },
+    {
+      "epoch": 0.8846451798872215,
+      "grad_norm": 4.13656759262085,
+      "learning_rate": 1.213960546282246e-06,
+      "loss": 0.2209,
+      "step": 3069
+    },
+    {
+      "epoch": 0.8855099356837877,
+      "grad_norm": 2.0688633918762207,
+      "learning_rate": 1.204855842185129e-06,
+      "loss": 0.2608,
+      "step": 3072
+    },
+    {
+      "epoch": 0.8863746914803539,
+      "grad_norm": 1.9340734481811523,
+      "learning_rate": 1.1957511380880123e-06,
+      "loss": 0.2187,
+      "step": 3075
+    },
+    {
+      "epoch": 0.8872394472769201,
+      "grad_norm": 2.4431509971618652,
+      "learning_rate": 1.1866464339908953e-06,
+      "loss": 0.2578,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8881042030734863,
+      "grad_norm": 2.9879822731018066,
+      "learning_rate": 1.1775417298937786e-06,
+      "loss": 0.2614,
+      "step": 3081
+    },
+    {
+      "epoch": 0.8889689588700525,
+      "grad_norm": 1.412150502204895,
+      "learning_rate": 1.1684370257966617e-06,
+      "loss": 0.1879,
+      "step": 3084
+    },
+    {
+      "epoch": 0.8898337146666186,
+      "grad_norm": 2.1106693744659424,
+      "learning_rate": 1.159332321699545e-06,
+      "loss": 0.2436,
+      "step": 3087
+    },
+    {
+      "epoch": 0.8906984704631848,
+      "grad_norm": 1.6913747787475586,
+      "learning_rate": 1.150227617602428e-06,
+      "loss": 0.2234,
+      "step": 3090
+    },
+    {
+      "epoch": 0.891563226259751,
+      "grad_norm": 1.6746766567230225,
+      "learning_rate": 1.1411229135053112e-06,
+      "loss": 0.2381,
+      "step": 3093
+    },
+    {
+      "epoch": 0.8924279820563172,
+      "grad_norm": 3.068824291229248,
+      "learning_rate": 1.1320182094081943e-06,
+      "loss": 0.223,
+      "step": 3096
+    },
+    {
+      "epoch": 0.8932927378528834,
+      "grad_norm": 2.9033825397491455,
+      "learning_rate": 1.1229135053110776e-06,
+      "loss": 0.2197,
+      "step": 3099
+    },
+    {
+      "epoch": 0.8935809897850722,
+      "eval_loss": 0.21448279917240143,
+      "eval_mse": 0.21448280355427415,
+      "eval_runtime": 6.6727,
+      "eval_samples_per_second": 149.865,
+      "eval_steps_per_second": 18.733,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8941574936494496,
+      "grad_norm": 1.8722037076950073,
+      "learning_rate": 1.1138088012139606e-06,
+      "loss": 0.1974,
+      "step": 3102
+    },
+    {
+      "epoch": 0.8950222494460158,
+      "grad_norm": 1.319684624671936,
+      "learning_rate": 1.1047040971168439e-06,
+      "loss": 0.2027,
+      "step": 3105
+    },
+    {
+      "epoch": 0.895887005242582,
+      "grad_norm": 3.9506242275238037,
+      "learning_rate": 1.095599393019727e-06,
+      "loss": 0.2124,
+      "step": 3108
+    },
+    {
+      "epoch": 0.8967517610391482,
+      "grad_norm": 1.7725896835327148,
+      "learning_rate": 1.08649468892261e-06,
+      "loss": 0.2146,
+      "step": 3111
+    },
+    {
+      "epoch": 0.8976165168357144,
+      "grad_norm": 1.9070608615875244,
+      "learning_rate": 1.0773899848254933e-06,
+      "loss": 0.2311,
+      "step": 3114
+    },
+    {
+      "epoch": 0.8984812726322806,
+      "grad_norm": 2.3098270893096924,
+      "learning_rate": 1.0682852807283763e-06,
+      "loss": 0.1847,
+      "step": 3117
+    },
+    {
+      "epoch": 0.8993460284288468,
+      "grad_norm": 2.392598867416382,
+      "learning_rate": 1.0591805766312596e-06,
+      "loss": 0.252,
+      "step": 3120
+    },
+    {
+      "epoch": 0.900210784225413,
+      "grad_norm": 1.592748761177063,
+      "learning_rate": 1.0500758725341426e-06,
+      "loss": 0.2226,
+      "step": 3123
+    },
+    {
+      "epoch": 0.9010755400219792,
+      "grad_norm": 2.1816020011901855,
+      "learning_rate": 1.0409711684370259e-06,
+      "loss": 0.2551,
+      "step": 3126
+    },
+    {
+      "epoch": 0.9019402958185454,
+      "grad_norm": 2.04571270942688,
+      "learning_rate": 1.031866464339909e-06,
+      "loss": 0.2506,
+      "step": 3129
+    },
+    {
+      "epoch": 0.9028050516151116,
+      "grad_norm": 3.148040771484375,
+      "learning_rate": 1.0227617602427922e-06,
+      "loss": 0.2056,
+      "step": 3132
+    },
+    {
+      "epoch": 0.9036698074116778,
+      "grad_norm": 4.721248626708984,
+      "learning_rate": 1.0136570561456753e-06,
+      "loss": 0.2243,
+      "step": 3135
+    },
+    {
+      "epoch": 0.904534563208244,
+      "grad_norm": 1.5059829950332642,
+      "learning_rate": 1.0045523520485585e-06,
+      "loss": 0.1884,
+      "step": 3138
+    },
+    {
+      "epoch": 0.9053993190048102,
+      "grad_norm": 1.9442839622497559,
+      "learning_rate": 9.954476479514416e-07,
+      "loss": 0.1965,
+      "step": 3141
+    },
+    {
+      "epoch": 0.9062640748013764,
+      "grad_norm": 2.1955509185791016,
+      "learning_rate": 9.863429438543249e-07,
+      "loss": 0.204,
+      "step": 3144
+    },
+    {
+      "epoch": 0.9071288305979426,
+      "grad_norm": 1.5504348278045654,
+      "learning_rate": 9.77238239757208e-07,
+      "loss": 0.2248,
+      "step": 3147
+    },
+    {
+      "epoch": 0.9079935863945088,
+      "grad_norm": 1.6235114336013794,
+      "learning_rate": 9.681335356600912e-07,
+      "loss": 0.2276,
+      "step": 3150
+    },
+    {
+      "epoch": 0.908858342191075,
+      "grad_norm": 2.8215818405151367,
+      "learning_rate": 9.590288315629742e-07,
+      "loss": 0.2292,
+      "step": 3153
+    },
+    {
+      "epoch": 0.9097230979876412,
+      "grad_norm": 1.7022403478622437,
+      "learning_rate": 9.499241274658574e-07,
+      "loss": 0.2383,
+      "step": 3156
+    },
+    {
+      "epoch": 0.9105878537842074,
+      "grad_norm": 2.5963640213012695,
+      "learning_rate": 9.408194233687407e-07,
+      "loss": 0.2311,
+      "step": 3159
+    },
+    {
+      "epoch": 0.9114526095807736,
+      "grad_norm": 1.830937147140503,
+      "learning_rate": 9.317147192716237e-07,
+      "loss": 0.2073,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9123173653773398,
+      "grad_norm": 1.8354508876800537,
+      "learning_rate": 9.22610015174507e-07,
+      "loss": 0.2143,
+      "step": 3165
+    },
+    {
+      "epoch": 0.913182121173906,
+      "grad_norm": 1.5569084882736206,
+      "learning_rate": 9.1350531107739e-07,
+      "loss": 0.2102,
+      "step": 3168
+    },
+    {
+      "epoch": 0.9140468769704722,
+      "grad_norm": 1.5762006044387817,
+      "learning_rate": 9.044006069802733e-07,
+      "loss": 0.2335,
+      "step": 3171
+    },
+    {
+      "epoch": 0.9149116327670384,
+      "grad_norm": 1.4155080318450928,
+      "learning_rate": 8.952959028831563e-07,
+      "loss": 0.1993,
+      "step": 3174
+    },
+    {
+      "epoch": 0.9157763885636045,
+      "grad_norm": 3.324310779571533,
+      "learning_rate": 8.861911987860396e-07,
+      "loss": 0.2209,
+      "step": 3177
+    },
+    {
+      "epoch": 0.9166411443601707,
+      "grad_norm": 2.4857962131500244,
+      "learning_rate": 8.770864946889227e-07,
+      "loss": 0.2384,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9175059001567369,
+      "grad_norm": 1.998383641242981,
+      "learning_rate": 8.679817905918058e-07,
+      "loss": 0.2216,
+      "step": 3183
+    },
+    {
+      "epoch": 0.9183706559533031,
+      "grad_norm": 2.2427866458892822,
+      "learning_rate": 8.58877086494689e-07,
+      "loss": 0.2328,
+      "step": 3186
+    },
+    {
+      "epoch": 0.9192354117498693,
+      "grad_norm": 2.5861330032348633,
+      "learning_rate": 8.497723823975721e-07,
+      "loss": 0.2228,
+      "step": 3189
+    },
+    {
+      "epoch": 0.9201001675464355,
+      "grad_norm": 1.680492639541626,
+      "learning_rate": 8.406676783004553e-07,
+      "loss": 0.2241,
+      "step": 3192
+    },
+    {
+      "epoch": 0.9209649233430017,
+      "grad_norm": 2.616267204284668,
+      "learning_rate": 8.315629742033385e-07,
+      "loss": 0.2524,
+      "step": 3195
+    },
+    {
+      "epoch": 0.921829679139568,
+      "grad_norm": 1.922303318977356,
+      "learning_rate": 8.224582701062215e-07,
+      "loss": 0.2397,
+      "step": 3198
+    },
+    {
+      "epoch": 0.9224061830039455,
+      "eval_loss": 0.21633096039295197,
+      "eval_mse": 0.2163309759118274,
+      "eval_runtime": 6.5496,
+      "eval_samples_per_second": 152.68,
+      "eval_steps_per_second": 19.085,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9226944349361342,
+      "grad_norm": 4.39849853515625,
+      "learning_rate": 8.133535660091048e-07,
+      "loss": 0.2216,
+      "step": 3201
+    },
+    {
+      "epoch": 0.9235591907327004,
+      "grad_norm": 2.2769124507904053,
+      "learning_rate": 8.042488619119878e-07,
+      "loss": 0.2182,
+      "step": 3204
+    },
+    {
+      "epoch": 0.9244239465292666,
+      "grad_norm": 2.8028557300567627,
+      "learning_rate": 7.951441578148711e-07,
+      "loss": 0.229,
+      "step": 3207
+    },
+    {
+      "epoch": 0.9252887023258328,
+      "grad_norm": 2.1702733039855957,
+      "learning_rate": 7.860394537177542e-07,
+      "loss": 0.2261,
+      "step": 3210
+    },
+    {
+      "epoch": 0.926153458122399,
+      "grad_norm": 2.6439995765686035,
+      "learning_rate": 7.769347496206374e-07,
+      "loss": 0.2246,
+      "step": 3213
+    },
+    {
+      "epoch": 0.9270182139189652,
+      "grad_norm": 1.576919436454773,
+      "learning_rate": 7.678300455235206e-07,
+      "loss": 0.2354,
+      "step": 3216
+    },
+    {
+      "epoch": 0.9278829697155314,
+      "grad_norm": 1.6755398511886597,
+      "learning_rate": 7.587253414264036e-07,
+      "loss": 0.2158,
+      "step": 3219
+    },
+    {
+      "epoch": 0.9287477255120976,
+      "grad_norm": 2.1890718936920166,
+      "learning_rate": 7.496206373292869e-07,
+      "loss": 0.2233,
+      "step": 3222
+    },
+    {
+      "epoch": 0.9296124813086638,
+      "grad_norm": 1.7316986322402954,
+      "learning_rate": 7.4051593323217e-07,
+      "loss": 0.2296,
+      "step": 3225
+    },
+    {
+      "epoch": 0.93047723710523,
+      "grad_norm": 1.7639137506484985,
+      "learning_rate": 7.314112291350532e-07,
+      "loss": 0.2274,
+      "step": 3228
+    },
+    {
+      "epoch": 0.9313419929017962,
+      "grad_norm": 1.7912460565567017,
+      "learning_rate": 7.223065250379363e-07,
+      "loss": 0.1942,
+      "step": 3231
+    },
+    {
+      "epoch": 0.9322067486983624,
+      "grad_norm": 1.9391751289367676,
+      "learning_rate": 7.132018209408196e-07,
+      "loss": 0.2326,
+      "step": 3234
+    },
+    {
+      "epoch": 0.9330715044949286,
+      "grad_norm": 2.4329934120178223,
+      "learning_rate": 7.040971168437026e-07,
+      "loss": 0.2205,
+      "step": 3237
+    },
+    {
+      "epoch": 0.9339362602914948,
+      "grad_norm": 1.5994445085525513,
+      "learning_rate": 6.949924127465859e-07,
+      "loss": 0.2149,
+      "step": 3240
+    },
+    {
+      "epoch": 0.934801016088061,
+      "grad_norm": 2.992966651916504,
+      "learning_rate": 6.858877086494689e-07,
+      "loss": 0.2194,
+      "step": 3243
+    },
+    {
+      "epoch": 0.9356657718846272,
+      "grad_norm": 1.8795028924942017,
+      "learning_rate": 6.767830045523521e-07,
+      "loss": 0.2158,
+      "step": 3246
+    },
+    {
+      "epoch": 0.9365305276811934,
+      "grad_norm": 3.5229902267456055,
+      "learning_rate": 6.676783004552352e-07,
+      "loss": 0.2396,
+      "step": 3249
+    },
+    {
+      "epoch": 0.9373952834777596,
+      "grad_norm": 1.6539433002471924,
+      "learning_rate": 6.585735963581184e-07,
+      "loss": 0.2077,
+      "step": 3252
+    },
+    {
+      "epoch": 0.9382600392743258,
+      "grad_norm": 2.448824405670166,
+      "learning_rate": 6.494688922610016e-07,
+      "loss": 0.2271,
+      "step": 3255
+    },
+    {
+      "epoch": 0.939124795070892,
+      "grad_norm": 1.3006185293197632,
+      "learning_rate": 6.403641881638847e-07,
+      "loss": 0.2432,
+      "step": 3258
+    },
+    {
+      "epoch": 0.9399895508674582,
+      "grad_norm": 2.689985752105713,
+      "learning_rate": 6.312594840667678e-07,
+      "loss": 0.2152,
+      "step": 3261
+    },
+    {
+      "epoch": 0.9408543066640244,
+      "grad_norm": 1.5370323657989502,
+      "learning_rate": 6.22154779969651e-07,
+      "loss": 0.197,
+      "step": 3264
+    },
+    {
+      "epoch": 0.9417190624605906,
+      "grad_norm": 1.6089318990707397,
+      "learning_rate": 6.130500758725342e-07,
+      "loss": 0.2145,
+      "step": 3267
+    },
+    {
+      "epoch": 0.9425838182571568,
+      "grad_norm": 2.1321656703948975,
+      "learning_rate": 6.039453717754174e-07,
+      "loss": 0.1989,
+      "step": 3270
+    },
+    {
+      "epoch": 0.943448574053723,
+      "grad_norm": 1.990070104598999,
+      "learning_rate": 5.948406676783005e-07,
+      "loss": 0.1982,
+      "step": 3273
+    },
+    {
+      "epoch": 0.9443133298502892,
+      "grad_norm": 3.087510585784912,
+      "learning_rate": 5.857359635811837e-07,
+      "loss": 0.206,
+      "step": 3276
+    },
+    {
+      "epoch": 0.9451780856468553,
+      "grad_norm": 2.3679909706115723,
+      "learning_rate": 5.766312594840668e-07,
+      "loss": 0.2292,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9460428414434215,
+      "grad_norm": 1.546036958694458,
+      "learning_rate": 5.675265553869499e-07,
+      "loss": 0.2127,
+      "step": 3282
+    },
+    {
+      "epoch": 0.9469075972399877,
+      "grad_norm": 1.9500880241394043,
+      "learning_rate": 5.584218512898331e-07,
+      "loss": 0.2157,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9477723530365539,
+      "grad_norm": 1.9931670427322388,
+      "learning_rate": 5.493171471927162e-07,
+      "loss": 0.2151,
+      "step": 3288
+    },
+    {
+      "epoch": 0.9486371088331201,
+      "grad_norm": 1.7962517738342285,
+      "learning_rate": 5.402124430955994e-07,
+      "loss": 0.2035,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9495018646296863,
+      "grad_norm": 1.3835334777832031,
+      "learning_rate": 5.311077389984825e-07,
+      "loss": 0.2059,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9503666204262525,
+      "grad_norm": 3.157975912094116,
+      "learning_rate": 5.220030349013658e-07,
+      "loss": 0.2002,
+      "step": 3297
+    },
+    {
+      "epoch": 0.9512313762228187,
+      "grad_norm": 2.0881123542785645,
+      "learning_rate": 5.12898330804249e-07,
+      "loss": 0.2307,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9512313762228187,
+      "eval_loss": 0.21600358188152313,
+      "eval_mse": 0.21600358275626785,
+      "eval_runtime": 6.4833,
+      "eval_samples_per_second": 154.243,
+      "eval_steps_per_second": 19.28,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9520961320193849,
+      "grad_norm": 2.1528842449188232,
+      "learning_rate": 5.037936267071321e-07,
+      "loss": 0.2657,
+      "step": 3303
+    },
+    {
+      "epoch": 0.9529608878159511,
+      "grad_norm": 2.4048590660095215,
+      "learning_rate": 4.946889226100153e-07,
+      "loss": 0.2355,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9538256436125173,
+      "grad_norm": 1.866373896598816,
+      "learning_rate": 4.855842185128983e-07,
+      "loss": 0.1932,
+      "step": 3309
+    },
+    {
+      "epoch": 0.9546903994090835,
+      "grad_norm": 1.543273687362671,
+      "learning_rate": 4.7647951441578155e-07,
+      "loss": 0.2088,
+      "step": 3312
+    },
+    {
+      "epoch": 0.9555551552056497,
+      "grad_norm": 1.585208535194397,
+      "learning_rate": 4.673748103186647e-07,
+      "loss": 0.1945,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9564199110022159,
+      "grad_norm": 1.9510430097579956,
+      "learning_rate": 4.582701062215478e-07,
+      "loss": 0.1719,
+      "step": 3318
+    },
+    {
+      "epoch": 0.9572846667987821,
+      "grad_norm": 1.36229407787323,
+      "learning_rate": 4.49165402124431e-07,
+      "loss": 0.1997,
+      "step": 3321
+    },
+    {
+      "epoch": 0.9581494225953483,
+      "grad_norm": 2.563950777053833,
+      "learning_rate": 4.4006069802731414e-07,
+      "loss": 0.2385,
+      "step": 3324
+    },
+    {
+      "epoch": 0.9590141783919145,
+      "grad_norm": 2.159186363220215,
+      "learning_rate": 4.309559939301973e-07,
+      "loss": 0.2319,
+      "step": 3327
+    },
+    {
+      "epoch": 0.9598789341884807,
+      "grad_norm": 2.4233345985412598,
+      "learning_rate": 4.2185128983308046e-07,
+      "loss": 0.2117,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9607436899850469,
+      "grad_norm": 1.915822982788086,
+      "learning_rate": 4.127465857359636e-07,
+      "loss": 0.2244,
+      "step": 3333
+    },
+    {
+      "epoch": 0.9616084457816131,
+      "grad_norm": 3.731882333755493,
+      "learning_rate": 4.0364188163884673e-07,
+      "loss": 0.2222,
+      "step": 3336
+    },
+    {
+      "epoch": 0.9624732015781793,
+      "grad_norm": 1.4122893810272217,
+      "learning_rate": 3.945371775417299e-07,
+      "loss": 0.2052,
+      "step": 3339
+    },
+    {
+      "epoch": 0.9633379573747455,
+      "grad_norm": 3.098508596420288,
+      "learning_rate": 3.8543247344461305e-07,
+      "loss": 0.212,
+      "step": 3342
+    },
+    {
+      "epoch": 0.9642027131713117,
+      "grad_norm": 1.4726747274398804,
+      "learning_rate": 3.763277693474962e-07,
+      "loss": 0.2374,
+      "step": 3345
+    },
+    {
+      "epoch": 0.9650674689678779,
+      "grad_norm": 1.8609498739242554,
+      "learning_rate": 3.6722306525037937e-07,
+      "loss": 0.2173,
+      "step": 3348
+    },
+    {
+      "epoch": 0.9659322247644441,
+      "grad_norm": 1.8475359678268433,
+      "learning_rate": 3.581183611532626e-07,
+      "loss": 0.2032,
+      "step": 3351
+    },
+    {
+      "epoch": 0.9667969805610104,
+      "grad_norm": 1.7029978036880493,
+      "learning_rate": 3.4901365705614574e-07,
+      "loss": 0.1713,
+      "step": 3354
+    },
+    {
+      "epoch": 0.9676617363575766,
+      "grad_norm": 2.657390832901001,
+      "learning_rate": 3.399089529590289e-07,
+      "loss": 0.2216,
+      "step": 3357
+    },
+    {
+      "epoch": 0.9685264921541428,
+      "grad_norm": 3.23854398727417,
+      "learning_rate": 3.3080424886191206e-07,
+      "loss": 0.1776,
+      "step": 3360
+    },
+    {
+      "epoch": 0.969391247950709,
+      "grad_norm": 2.09384822845459,
+      "learning_rate": 3.2169954476479517e-07,
+      "loss": 0.1866,
+      "step": 3363
+    },
+    {
+      "epoch": 0.9702560037472752,
+      "grad_norm": 1.8957816362380981,
+      "learning_rate": 3.1259484066767833e-07,
+      "loss": 0.2317,
+      "step": 3366
+    },
+    {
+      "epoch": 0.9711207595438414,
+      "grad_norm": 1.5353327989578247,
+      "learning_rate": 3.034901365705615e-07,
+      "loss": 0.209,
+      "step": 3369
+    },
+    {
+      "epoch": 0.9719855153404076,
+      "grad_norm": 1.776352882385254,
+      "learning_rate": 2.9438543247344465e-07,
+      "loss": 0.219,
+      "step": 3372
+    },
+    {
+      "epoch": 0.9728502711369738,
+      "grad_norm": 3.282552480697632,
+      "learning_rate": 2.852807283763278e-07,
+      "loss": 0.224,
+      "step": 3375
+    },
+    {
+      "epoch": 0.97371502693354,
+      "grad_norm": 1.8530248403549194,
+      "learning_rate": 2.7617602427921097e-07,
+      "loss": 0.2147,
+      "step": 3378
+    },
+    {
+      "epoch": 0.9745797827301061,
+      "grad_norm": 2.0258686542510986,
+      "learning_rate": 2.670713201820941e-07,
+      "loss": 0.2081,
+      "step": 3381
+    },
+    {
+      "epoch": 0.9754445385266723,
+      "grad_norm": 1.830100655555725,
+      "learning_rate": 2.5796661608497724e-07,
+      "loss": 0.2248,
+      "step": 3384
+    },
+    {
+      "epoch": 0.9763092943232385,
+      "grad_norm": 3.846280813217163,
+      "learning_rate": 2.488619119878604e-07,
+      "loss": 0.2269,
+      "step": 3387
+    },
+    {
+      "epoch": 0.9771740501198047,
+      "grad_norm": 2.3556368350982666,
+      "learning_rate": 2.3975720789074356e-07,
+      "loss": 0.2195,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9780388059163709,
+      "grad_norm": 1.8791780471801758,
+      "learning_rate": 2.3065250379362674e-07,
+      "loss": 0.2297,
+      "step": 3393
+    },
+    {
+      "epoch": 0.9789035617129371,
+      "grad_norm": 1.6351841688156128,
+      "learning_rate": 2.215477996965099e-07,
+      "loss": 0.1925,
+      "step": 3396
+    },
+    {
+      "epoch": 0.9797683175095033,
+      "grad_norm": 2.4028263092041016,
+      "learning_rate": 2.1244309559939304e-07,
+      "loss": 0.2099,
+      "step": 3399
+    },
+    {
+      "epoch": 0.9800565694416921,
+      "eval_loss": 0.2100730687379837,
+      "eval_mse": 0.21007308381050824,
+      "eval_runtime": 6.746,
+      "eval_samples_per_second": 148.235,
+      "eval_steps_per_second": 18.529,
+      "step": 3400
+    },
+    {
+      "epoch": 0.9806330733060695,
+      "grad_norm": 3.0006494522094727,
+      "learning_rate": 2.033383915022762e-07,
+      "loss": 0.2095,
+      "step": 3402
+    },
+    {
+      "epoch": 0.9814978291026357,
+      "grad_norm": 2.7023770809173584,
+      "learning_rate": 1.9423368740515936e-07,
+      "loss": 0.2064,
+      "step": 3405
+    },
+    {
+      "epoch": 0.9823625848992019,
+      "grad_norm": 3.8468315601348877,
+      "learning_rate": 1.851289833080425e-07,
+      "loss": 0.2412,
+      "step": 3408
+    },
+    {
+      "epoch": 0.9832273406957681,
+      "grad_norm": 2.5841078758239746,
+      "learning_rate": 1.7602427921092565e-07,
+      "loss": 0.2056,
+      "step": 3411
+    },
+    {
+      "epoch": 0.9840920964923343,
+      "grad_norm": 2.557180881500244,
+      "learning_rate": 1.669195751138088e-07,
+      "loss": 0.2449,
+      "step": 3414
+    },
+    {
+      "epoch": 0.9849568522889005,
+      "grad_norm": 2.1071228981018066,
+      "learning_rate": 1.5781487101669194e-07,
+      "loss": 0.2664,
+      "step": 3417
+    },
+    {
+      "epoch": 0.9858216080854667,
+      "grad_norm": 2.5804591178894043,
+      "learning_rate": 1.4871016691957513e-07,
+      "loss": 0.2252,
+      "step": 3420
+    },
+    {
+      "epoch": 0.9866863638820329,
+      "grad_norm": 1.557554006576538,
+      "learning_rate": 1.3960546282245826e-07,
+      "loss": 0.1979,
+      "step": 3423
+    },
+    {
+      "epoch": 0.9875511196785991,
+      "grad_norm": 2.0957911014556885,
+      "learning_rate": 1.3050075872534145e-07,
+      "loss": 0.21,
+      "step": 3426
+    },
+    {
+      "epoch": 0.9884158754751653,
+      "grad_norm": 2.239748954772949,
+      "learning_rate": 1.2139605462822459e-07,
+      "loss": 0.1995,
+      "step": 3429
+    },
+    {
+      "epoch": 0.9892806312717315,
+      "grad_norm": 2.46882963180542,
+      "learning_rate": 1.1229135053110775e-07,
+      "loss": 0.2319,
+      "step": 3432
+    },
+    {
+      "epoch": 0.9901453870682977,
+      "grad_norm": 2.6145310401916504,
+      "learning_rate": 1.031866464339909e-07,
+      "loss": 0.2245,
+      "step": 3435
+    },
+    {
+      "epoch": 0.9910101428648639,
+      "grad_norm": 2.32053279876709,
+      "learning_rate": 9.408194233687405e-08,
+      "loss": 0.2338,
+      "step": 3438
+    },
+    {
+      "epoch": 0.9918748986614301,
+      "grad_norm": 1.4136054515838623,
+      "learning_rate": 8.497723823975723e-08,
+      "loss": 0.2173,
+      "step": 3441
+    },
+    {
+      "epoch": 0.9927396544579963,
+      "grad_norm": 1.3618991374969482,
+      "learning_rate": 7.587253414264037e-08,
+      "loss": 0.1794,
+      "step": 3444
+    },
+    {
+      "epoch": 0.9936044102545625,
+      "grad_norm": 2.8345561027526855,
+      "learning_rate": 6.676783004552352e-08,
+      "loss": 0.2182,
+      "step": 3447
+    },
+    {
+      "epoch": 0.9944691660511287,
+      "grad_norm": 1.9578803777694702,
+      "learning_rate": 5.7663125948406686e-08,
+      "loss": 0.2333,
+      "step": 3450
+    },
+    {
+      "epoch": 0.9953339218476949,
+      "grad_norm": 1.6162538528442383,
+      "learning_rate": 4.855842185128984e-08,
+      "loss": 0.2005,
+      "step": 3453
+    },
+    {
+      "epoch": 0.9961986776442611,
+      "grad_norm": 3.075516939163208,
+      "learning_rate": 3.9453717754172986e-08,
+      "loss": 0.2249,
+      "step": 3456
+    },
+    {
+      "epoch": 0.9970634334408273,
+      "grad_norm": 2.544214963912964,
+      "learning_rate": 3.0349013657056146e-08,
+      "loss": 0.2147,
+      "step": 3459
+    },
+    {
+      "epoch": 0.9979281892373935,
+      "grad_norm": 1.9591963291168213,
+      "learning_rate": 2.1244309559939306e-08,
+      "loss": 0.2164,
+      "step": 3462
+    },
+    {
+      "epoch": 0.9987929450339597,
+      "grad_norm": 1.6480742692947388,
+      "learning_rate": 1.213960546282246e-08,
+      "loss": 0.1889,
+      "step": 3465
+    },
+    {
+      "epoch": 0.9996577008305259,
+      "grad_norm": 1.6390196084976196,
+      "learning_rate": 3.034901365705615e-09,
+      "loss": 0.2205,
+      "step": 3468
+    },
+    {
+      "epoch": 0.9999459527627146,
+      "step": 3469,
+      "total_flos": 1.1682867916662374e+17,
+      "train_loss": 0.2771636627187094,
+      "train_runtime": 4603.777,
+      "train_samples_per_second": 96.454,
+      "train_steps_per_second": 0.754
+    }
+  ],
+  "logging_steps": 3,
+  "max_steps": 3469,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1682867916662374e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}